Tidal rectification and exchange at the neck of a canyon–bay system: a tides-only modeling framework

Submarine canyons can convert oscillatory tidal forcing into a time-mean flow via nonlinear rectification and wave–mean interactions, with consequences for exchange across narrow constrictions. Here, we present a tides-only numerical study focusing on residual transport at the neck of the Punta Banda submarine canyon, which connects the southern entrance of Todos Santos Bay (Ensenada, Baja California, Mexico) with the Pacific Ocean. The canyon contracts from 3 km (mouth) to 2 km (neck) and shoals from ~700 m to ~400 m, giving an inner-to-mouth cross-section ratio of ~1.5—a geometry favorable to flood–ebb asymmetry, frictional losses, and form drag even for modest currents. However, for the dominant constituent M2, the depth-mean speed is O(0.01 m s⁻¹), giving a Keulegan–Carpenter number KC ≈ 0.3 at the neck and a Rossby number Ro ≈ 0.1, suggesting that nonlinear effects and rectification might be relatively weak. Stratification is seasonally strong above the rim, and we estimate a Burger number Bu ≈ 1.0 near ~150 m (canyon rim depth), consistent with efficient baroclinic waveguiding. To test the dominant regime, we use the MIT general circulation model (MITgcm) forced at the open boundaries with the dominant M2 and K1 barotropic constituents from TPXO10, and realistic but horizontally homogeneous temperature and salinity profiles. We map tidal ellipses, diagnose Eulerian residuals, calculate residual volume transport to quantify rectification strength, and close the time-mean momentum budget across the neck, separating pressure gradients, Reynolds stresses, bottom drag, and form drag. To test physical mechanisms, we compare runs with and without stratification, nonlinear advection, and varying grid size. Together, these experiments assess whether tides alone can drive persistent along-canyon mean flow and net exchange at the Punta Banda canyon neck.