Baroclinic Forcing and Vertical Shear in the Ligurian Current
The coastal waters off Sanremo present a brutal environment for current profiling because of the extreme vertical shear. I've seen profiles where the surface layer is sprinting east at 0.3 m/s while the water just 50 meters down is pushing west. This isn't a simple linear gradient. It's a volatile, two-layered system driven by the interaction between the dense, salty Ligurian Current and the atmospheric forcing of the Mistral winds. When those northwest winds hit, they shove surface waters toward the coast, creating a setup that overrides the regional flow. If you're using surface drifters here, you're essentially guessing. You're seeing the skin of the ocean, not the engine.
Most people mistake the microtidal nature of the Gulf of Genoa for stability. The tidal range here stays under 30cm, but that's a distraction. The real energy comes from the baroclinic pressure gradients. The water column is rarely in equilibrium. We often observe intense turbulence where the wind-driven surface layer rubs against the deeper, colder Mediterranean flow. This friction creates eddies that can throw off a low-frequency sensor. You get these erratic spikes in your data that look like instrument failure but are actually just the chaotic reality of the Sanremo shelf.
The density stratification here is particularly aggressive during the transition from spring to summer. As the surface warms, the pycnocline sharpens. This creates a physical boundary that doesn't just affect biology; it bends acoustic energy. In my experience, ignoring the temperature-salinity profile in this region leads to significant errors in sound speed correction. If you don't account for the local speed of sound, your depth bins shift. Your data becomes a lie.
The Steep Bathymetry of the Sanremo Shelf
Sanremo sits on a geological knife-edge. Unlike the broad, sweeping banks of the North Sea, the continental shelf here is practically non-existent. Within a few nautical miles of the shoreline, the seabed plunges. You can be in 20 meters of water and, with a slight shift in position, find yourself over a 500-meter drop-off. This rugged topography acts as a funnel for the Ligurian Current. The current doesn't just flow past; it interacts with the seabed, creating localized accelerations and wake effects that make ground-truthing a nightmare.
Looking at the contours around the Sanremo coast (roughly 43.6°N, 7.6°E), the gradient is staggering. The current is squeezed between the coast and the deep basin. This creates a venturi effect. In some narrow corridors, the flow velocity spikes unexpectedly. I've found that placing a sensor too close to these steep slopes leads to 'noisy data' because of the turbulence generated by the bottom boundary layer. You need to position your mooring carefully or you'll spend weeks filtering out signal noise caused by seabed friction.
Acoustic Propagation Challenges in This Environment
The Mediterranean is salty. The Ligurian Sea is even saltier. This high salinity, combined with rapid seasonal heating, makes acoustic propagation unpredictable. The biggest headache is the summer thermocline. When the surface layer hits 22°C while the depths remain at 13°C, you get a refractive boundary. This can cause 'shadow zones.' The acoustic pings from an ADCP can actually bend away from the target area. I've seen cases where the signal attenuation is so high that the sensor simply can't 'see' the water column clearly, leaving you with gaps in your profile precisely where the most interesting shear is happening.
Then there's the winter runoff. When the Alps melt or heavy rains hit, the runoff into the Ligurian Sea increases turbidity. While we aren't dealing with the mud of a river delta, the suspended particulate matter increases. This causes signal scattering. If your blanking distance is set too short, you'll get bin contamination. I remember a deployment near Genoa where storm-driven whitecaps created so much surface noise that the top three bins were completely wiped out. We had to recalibrate the signal fence on the fly to salvage the rest of the data. It's a constant battle between wanting high resolution and needing a clean signal.
Frequency Selection and Deployment Logic
For this specific environment, I strongly argue against using low-frequency ADCPs. You need the resolution. A 600kHz or even a 1200kHz unit is the only way to capture the sharp vertical gradients characteristic of the Sanremo coast. The 600kHz unit generally outperforms others here because it balances the need for a decent range with the precision required to distinguish between the wind-driven layer and the deeper flow. If you go too low in frequency, you blur the shear. You end up with an average velocity that doesn't actually exist at any single depth. That's useless for actual hydrodynamic modeling.
Deployment must be bottom-mounted and rigidly fixed. Given the steep slopes, a tripod mount with a heavy ballast is mandatory. I've seen moorings slide down the bathymetric gradient during a Mistral event, which turns your time-series data into a guessing game. You must use a high-precision GPS for the initial drop and then perform a sanity check with a secondary acoustic release. If the instrument tilts more than a few degrees, your tilt-correction algorithms will struggle, and you'll start seeing phantom currents.
Data Interpretation and Field Findings
When we analyze the data from Sanremo, the first thing we look for is the 'cross-over' depth. This is the point where the velocity vector flips from east to west. In a typical autumn profile, this flip can happen within a 15-meter window. If the data shows a smooth transition, I suspect the sampling interval is too wide. Real Sanremo data is jagged. It's violent. You see these sudden bursts of velocity that correlate exactly with wind gusts recorded at the shore. The correlation is almost 1:1 for the top 20 meters.
We also see significant 'aliasing' if the averaging period is too short. If you set your ADCP to average over 15 minutes, you might miss the peak pulse of a tidal surge or a wind-driven event. However, if you go too short, the signal-to-noise ratio drops. I've found that a 30-minute ensemble is the sweet spot for this region. It smooths out the random turbulence but keeps the signal sharp enough to identify the onset of a Mistral-driven flow. Anything else is just playing with noise.
Operational Implications
These current patterns aren't just academic. They have massive implications for local maritime operations. For cable laying or underwater sensor deployment near Sanremo, the vertical shear means that a cable can be under immense tension at the surface while being slack at the bottom. It's a recipe for fatigue failure. Engineers who assume a uniform current profile across the water column are asking for trouble. You cannot treat the Ligurian coast as a monolithic block of moving water.
Furthermore, for port management and coastal erosion studies, understanding the 'setup' created by the Mistral is critical. The way the surface water piles up against the coast changes the sediment transport patterns. By using high-resolution acoustic profiling, we can finally quantify how much sediment is being stripped from the nearshore zone during winter storms. It's the difference between guessing and actually knowing the volumetric transport. Without a vertical profile, you're just looking at a snapshot of a much larger, more complex machine.
About the author: Sarah Jenkins. A world-class expert in underwater acoustics and oceanographic instrumentation specializing in tidal asymmetry and continental shelf currents. She has spent two decades deploying sensors in the world's most challenging bathymetric environments.
Resolving Vertical Velocity Shear and Acoustic Refraction in the Sanremo Coastal Zone