Bora-Induced Velocity Spikes and Vertical Shear in the Southern Adriatic
Field observations at the tip of the Istrian Peninsula frequently reveal surface velocities exceeding 1.2 m/s during peak Bora events, while bottom-track data often remains stagnant or reverses direction entirely. This extreme vertical shear is the defining characteristic of Pula's coastal waters. Generic hydrodynamic models fail here because they treat the water column as a cohesive unit. In reality, the Bora—a violent, cold northeasterly wind—creates a decoupled surface layer that slides over the denser, slower-moving Adriatic waters. This doesn't just complicate the math; it creates a nightmare for acoustic profiling.
The interaction between these wind-driven surges and the rugged coastline generates localized turbulence that shreds the coherence of acoustic backscatter. I've seen data sets where the upper 5 meters of the water column show massive velocity oscillations while the lower 20 meters are dead quiet. This creates a high-gradient shear zone that leads to severe bin contamination. If your bin size is too large, the high-velocity surface signal bleeds into the lower cells, giving you an average velocity that doesn't exist at any actual depth. It's a classic case of spatial averaging masking the real physics of the site.
Tidal signals in the Adriatic are modest, but in Pula, they get completely drowned out by these atmospheric forcing events. We aren't dealing with a simple ebb and flow. Instead, we see a chaotic superposition of wind-driven currents, density-driven flows, and the general cyclonic circulation of the Adriatic. This makes ground-truthing incredibly difficult. You can't just rely on a tide gauge to understand what's happening 10 meters down.
The Bathymetric Complexity of the Istrian Peninsula Tip
Pula sits at a geographic pinch point where the seabed transitions violently from shallow limestone reef flats to deep underwater canyons. Around coordinates 44.5°N, 13.8°E, the bathymetry is a chaotic mess of jagged karst topography. These deep channels act as conduits for water masses moving along the eastern Adriatic coast. When the Bora pushes surface water away from the coast, these canyons often facilitate a compensatory inflow of deeper, saltier water. This creates a complex three-dimensional flow pattern that defies linear prediction.
The coastline is heavily indented, which focuses the energy of the Jugo (Southeasterly) and Bora winds into narrow corridors. This focusing effect amplifies current speeds in specific channels while leaving adjacent bays in a state of relative stagnation. I've noticed that the eddies formed at the edges of these channels are particularly aggressive. They create rotational flow that can trick a bottom-mounted ADCP into recording a current that is actually a localized vortex rather than a regional flow trend. You have to be obsessive about your site selection to avoid these 'dead zones' or 'hot spots' that skew the data.
Acoustic Propagation Challenges in This Environment
The Adriatic is notorious for its salinity gradients, but Pula adds a layer of complexity with its high turbidity during storm events. The Bora doesn't just move the water; it scours the seabed in the shallower bays, kicking up organic debris and fine sediments. This suspended load increases the attenuation of the acoustic signal. In my experience, this leads to 'noisy data' in the upper bins, where the signal-to-noise ratio drops precipitously. The backscatter becomes erratic, and the correlation peak becomes difficult to identify, leading to frequent 'bad data' flags in the raw files.
Temperature inversions also play a role. During winter Bora events, the surface water cools rapidly, creating a sharp thermocline. This density interface can act as a partial reflector for acoustic energy or cause refraction of the sonar beams. While not as severe as the deep-ocean SOFAR channel, these local gradients can introduce subtle errors in the sound speed profile. If you use a constant sound speed instead of updating the profile with CTD casts, your depth calculations will be off. It sounds minor, but in a 30-meter water column, a 1% error in sound speed can shift your bins enough to misidentify the depth of the shear layer.
Frequency Selection and Deployment Strategy
Choosing between 300kHz and 600kHz for Pula depends entirely on the specific deployment site. In the deeper channels off the coast, 300kHz is the only logical choice to get the range needed to see the bottom. However, for the sheltered bays and reef areas, I always insist on 600kHz. The reason is simple: blanking distance. A 300kHz unit has a larger blanking distance, meaning you lose the first few meters of data. In Pula, the Bora's most critical dynamics happen in those first 2-5 meters. Losing that data is like trying to understand a storm while wearing a blindfold.
Mooring strategy is where most people mess up in the Adriatic. Vessel-mounted units are useless during high-wind events because the surface chop is too violent. Bottom-mounted frames with heavy ballast are the only way to ensure stability. I've seen lightweight tripods tilt 15 degrees during a Bora surge, which completely ruins the geometric correction of the ADCP. You need a frame that is effectively bolted to the seabed. Honestly, the more ballast you can throw at it, the better. A stable platform is the only way to ensure that the measured velocity is actually water movement and not the instrument swaying in the current.
Data Interpretation and Field Findings
When reviewing the data from Pula, the first thing I look for is the correlation magnitude. If the correlation drops during a velocity spike, I know I'm looking at sediment-induced noise rather than a real current. We've found that during peak Bora events, the vertical shear is so intense that the 'effective' current is a fiction. You might see 0.9 m/s at the surface and 0.1 m/s at 10 meters. Averaging these gives you 0.5 m/s, but that number is useless for hydrodynamic modeling because it describes a state that doesn't exist at any point in the water column.
The most interesting finding in this region is the lag time between wind peak and current peak. The surface responds almost instantly to the Bora, but the deeper layers respond with a delay or not at all. This hysteresis loop is a key indicator of the local stability of the water column. In some deployments, we've seen the bottom layers actually accelerate in the opposite direction of the surface flow (a counter-current) as the system attempts to maintain mass balance. This confirms that Pula's coastal zone acts more like a complex estuary than a simple open coast.
Operational Implications
These hydrodynamic anomalies have real-world consequences for maritime operations in Pula. For anyone deploying underwater sensors or conducting dredging, the Bora-driven shear means that equipment can be subjected to asymmetric loads. A cable that is slack at the bottom might be under extreme tension at the surface. This leads to premature fatigue and unexpected mooring failures.
From a monitoring perspective, the high volatility means that short-term deployments (e.g., 2 weeks) are almost useless. You might hit a period of total calm or a series of violent storms, giving you a skewed view of the annual mean. To get a sanity check on the data, you need long-term deployments that span multiple seasons. Only then can you differentiate between a seasonal trend and a random wind event. For those managing port infrastructure, understanding these shear profiles is critical for predicting sediment transport and siltation rates in the harbor.
About the author: Sarah Jenkins. Sarah is a leading expert in underwater acoustics and oceanographic instrumentation with twenty years of experience in continental shelf dynamics. She specializes in the application of ADCP technology in high-shear coastal environments.
Mitigating Vertical Shear and Acoustic Noise during Bora Wind Events in the Pula Coastal Zone