No abstract

The authors describe a two-dimensional (vector) regressional model for examining the anisotropic response of ice drift and ocean current velocity (“drift velocity”) to surface wind forcing. Illustration of the method is limited to sea ice response. The principal mathematical and physical properties of the model are outlined, together with estimates of the “response matrices” and the corresponding “response ellipses” (drift velocity response to a unity wind velocity forcing). For each direction, φ, of the wind vector the method describes a corresponding “wind factor” α(φ) (relative drift speed) and “turning angle” θ(φ) (the angle between the drift velocity and wind vector). The major ellipse axis corresponds to the direction of the “effective wind” (φ = φmax) and the minor axis to the direction of the “noneffective” wind. The eigenvectors of the response matrix define wind directions that are the same as the wind-induced drift velocity directions. Depending on the water depth and offshore distance, six analytical cases are possible, ranging from rectilinear response ellipses near the coast to purely circular response ellipses in the open ocean. The model is used to examine ice drift along the western shelf of Sakhalin Island (Sea of Okhotsk). Responses derived from the vector regression (four parameter) method are less constrained and therefore more representative of wind-induced surface motions than those derived using the traditional complex transfer function (two parameter) approach.

Ice-drift velocity records from coastal radar stations, combined with data from moored current meters and coastal wind stations, are used to examine sea-ice motion off the northeastern coast of Sakhalin Island in the Sea of Okhotsk. Ice motion is shown to be governed primarily by diurnal tidal currents and wind-induced drift, which explain 92%–95% of the total ice-drift variance. Diurnal tidal motions predominate off the northern Sakhalin coast, accounting for 65%–80% of the variance, while low-frequency wind-induced motions prevail off the south-central coast, accounting for over 91% of the ice-drift variance. Maximum diurnal tidal ice-drift velocities range from 90–110 cm s−1 on the north coast to 10–15 cm s−1 on the south coast, in good agreement with the barotropic model of Kowalik and Polyakov. The presence of diurnal shelf waves accounts for the strong diurnal currents on the steeply sloping northern Sakhalin shelf, while the absence of such waves explains the weak diurnal currents on the more gently sloping south-central shelf. Using a vector regression model, the authors show that wind-induced ice-drift “response ellipses” (the current velocity response to a unity wind-velocity forcing) are consistent with a predominantly alongshore response to the wind, with wind-induced currents most pronounced off the south-central coast where water depths are relatively shallow. Time–frequency analysis of wind and ice-drift series reveals that, in winter, when sea ice is most extensive and internally cohesive, the ice response is almost entirely aligned with the alongshore component of the wind; in spring, when sea ice is broken and patchy, the ice responds to both the cross- and alongshore components of the wind.