S-PIV boundary layer set-up configuration.
Radial profile of the azimuthal velocity for spin-up and spin-down cases (note the calculation area for interior velocity).
Snapshots taken during the SD30-60-H66 spin-down case. (Left) Top view of instabilities in the Ekman layer developing during the initiation of the spin-down flow at t = 2 s after the flow initiation. (Right) Top view of the fully turbulent Ekman layer at t = 42 s after the flow initiation.
Time series of Eulerian interior velocity for spin-up case SU∞-60-H66 and spin-down case SD30-60-H66. The two parallel dashed lines represent the predicted exponential Ekman decay for laminar regime (assumed for U < 5 cm/s).
Evolution of the dimensionless velocity U/u ⋆ versus the square root of the friction Rossby number for smooth bottom for the spin-up SU∞-60-H66 and spin-down SD30-60-H66 cases. Present data are compared with experimental data of Caldwell et al. 4 (with modified Coriolis parameter) and ABL theory for different values of universal constants A and B: best estimates from atmospheric measurements 12 (A = 1.3, B = 4.4) and best fit for present experiments (A = 3.3, B = 3).
Evolution of the dimensionless velocity U/u ⋆ vs the square root of the friction Rossby number for smooth bottom . For each plot, the reference curve of Fig. 5 is added for comparison (laser height 50 cm). (Left plot) Dependance of the vertical position with four selected heights z = 1, 5, 30, 50 cm in the interior flow for the spin-up case SU60-30-H66. (Right plot) Influence of water depth with three different water depths H = 66, 30, and 15 cm for the spin-up cases SU∞-60-H66, SU∞-60-H30, and SU∞-60-H15, respectively.
Comparison between experimental velocity profile (dashed thick line for the radial component, dashed thin line for the azimuthal component) and Ekman theoretical solutions (solid thick line for the radial component, solid thin line for the azimuthal component) for boundary layer Reynolds number Re δ = 45 (t = 4750 s).
Time evolution of the vertical profiles of the three components of the velocity from Re δ = 360 to Re δ = 88 corresponding to t = 562 and 3390 s, respectively.
Comparison between and , with z g = 3 cm.
Evolution of the cross-isobar angle α0 vs the square root of the friction Rossby number for smooth bottom . (Left) Present data only. (Right) Comparison of these data with ABL theory and Caldwell et al. laboratory measurements. 3
(Left) Time evolution of the Ekman spiral from t = 583 to 4580 s (radial vs azimuthal velocity components). (Right top) Normalized velocity defect parallel to surface stress versus normalized height using the outer scaling to the similarity theories, comparison with Caldwell et al. 4 experimental data. (Right bottom) Normalized velocity defect normal to surface stress versus normalized height using the outer scaling to the similarity theories, comparison with Caldwell et al. 4 experimental data.
Probability density functions for the normalized vertical velocity at heights z = 1, 2, 3, 5 cm for the laminar (Re δ = 68, t = 3920 s) and turbulent (Re δ = 250, t = 1290 s) regimes.
(Left) Normalized shear stress in direction of surface stress versus height normalized by the outer scaling. (Right) Normalized shear stress normal to direction of surface stress versus height normalized by the outer scaling. Turbulence measurements below z = 0.3 cm are not reliable due to spurious light diffusion near the bottom.
List of experiments with values of control parameters with T ini and T the initial and final rotation periods, respectively, H the water depth, and Ω the angular rotation rate.
Article metrics loading...
Full text loading...