The subtidal behaviour of the Celtic Sea—I. Sea level and bottom pressures

Interest in the subtidal behaviour of shelf seas calls for coordinated programmes of long period and shorter period intensive measurements, together with analytical and numerical representation of their responses to forcing. The following three papers consider related aspects of the behaviour of the Celtic Sea. This first account sets the background and considers the evidence in the sea level and bottom pressure records for responses to local and external forcing. The second paper considers the subtidal currents in terms of momentum balance, with special reference to the effects of clockwise and counterclockwise wind forcing, and the influence of external pressure gradients. The third paper is a related theoretical study of shelf waves on a wide shelf, which takes as an example the observed responses of the Celtic Sea. Analysis of more than 30 years of simultaneous sea levels and air pressures shows that at Newlyn observed levels do not follow the inverted barometer effect at any frequency. At frequencies higher than 0.1 cpd the levels anticipate an inverted barometer response because of a lagged correlation between northward winds and low atmospheric pressures: as a result there is a pronounced change of gradient in the energy spectrum below 0.1 cpd which is not seen in the atmospheric pressures. It is suggested that this is due to the dominance of eastward-propagating weather patterns above 0.1 cpd, compared with the symmetric propagation properties of weather patterns below 0.1 cpd ( WILLEBRAND, 1978, Journal of Physical Oceanography, 8, 1080-1094). The Celtic Sea has an essentially uniform response to atmospheric forcing, which conforms to the theoretical analysis in Part III, for idealized ocean-shelf configurations. The year and the month of detailed data show that for both sea levels and total pressure, 90% of the variability is accounted for by the first principal component. The response to this component is weakest at the shelf edge and strongest at the entrance to the Bristol Channel. After removing variations which are coherent over the whole sea, coastal levels have a residual signal with a 2 to 3 cm standard deviation; in contrast the best offshore pressure records have a residual of < 1mb. This enhanced noise level for coastal measurements, which is probably due to non-linear wave effects and nearshore currents, defines a limit to the use of coastal levels for monitoring regional dynamics. When the Newlyn levels are compared with the levels at an offshelf ocean station, both being adjusted for meteorologically correlated variations, there is a weak but significant correlation. The coherence is strongest at 0.25 cpd, when ocean-bed pressures lead Newlyn; however, the length of simultaneous data is too short for firm conclusions about the ocean pressures driving the shelf.