Modelling and measuring the wind forced inflow to the Irish Sea through the North Channel

摘要

The wind-induced flow though the North Channel of the Irish Sea during the period 13 March-26 March 1994 has been investigated using a range of three-dimensional models, and current observations were made using an acoustic Doppler current profiler (ADCP). Computed surface elevations are compared with shore based and one off shore tide gauge measurements. A shelf wide coarse grid (of order 12 km resolution) model forced by hourly varying wind stresses and pressure gradients, is used to provide boundary conditions for a finer grid (of order 5 km) west coast model. A high-resolution model (of order 1 km) is used to resolve the North Channel region. Calculations with the large area models show the importance of shelf wide wind events particularly those along the west coast of Britain in influencing the flow through the North Channel. Two different meteorological situations producing a flow to the south in the North Channel (an inflow to the Irish Sea) are identified during the simulations. The first is a region of strong winds from the west, off the west coast of Scotland, with only light winds off Ireland and over the Irish Sea, These winds force a flow to the south along the west coast of Scotland, through the North Channel and into the Irish Sea. The second meteorological situation is a region of uniform winds from the southwest over the whole west coast of Britain, which then changes to a wind field from the west. This drives a flow to the north along the west coast of Ireland, part of which continues to the north, although some of the flow passes round northern Ireland, and then flows to the south within the North Channel. The response of the various models to these changing wind events are studied in detail and computed currents are compared with those measured using the ADCP. This shows that computed along-channel currents are underpredicted at all depths, although there is no bias in the computed across-channel currents away from the near-bed layer. However, close to the bed the model has a bias to overpredict the across channel flow. These biases are probably due to inaccuracies in resolving the topography of the deep water region in the central North Channel. On average the model reproduces the current to within +/-4 cm s(-1) for the majority of the period. Similarly, sea surface elevations are accurately (to within 12 cm RMS error) reproduced. (C) 2002 Elsevier Science Ltd. All rights reserved. [References: 46]