Subsidence of Rockall Plateau and of the Continental Shelf

Matthews & Smith (1971) have rightly studied the problem of the sinking of the Rockall-Hatton sedimentary basin in terms of two possible hypotheses, both involving isostatic subsidence in response to crustal thinning. These are (1) crustal thinning through inversion of a gabbroic lower crustal layer to eclogite (attributed to Collette), and (2) crustal thinning by seaward creep of the lower part (attributed to Bott). In working out the implications of the latter hypothesis, Matthews & Smith deal with a substantially different mechanism from one which I briefly put forward at the end of a paper to the SCOR symposium on the Eastern Atlantic continental margin (Bott & Watts 1971). An account of my hypothesis has been published (Bott 1971), but this was presumably not available when Matthews & Smith submitted their paper. According to Matthews & Smith, extension of the upper crust by normal faulting is identical to the extension of the lower crust by creep. Crustal thinning can then be estimated from the stretching of the upper crust produced by normal faulting. Matthews & Smith show that a total throw of 40 km is required to account for the subsidence of the Rockall-Hatton basin by this mechanism. The seismic reflection results convince me that normal faulting does not occur on the scale required by the MattheweSmith version of the creep hypothesis. This mechanism has not been seriously suggested for shelf subsidence and basin formation simply because normal faulting does not appear to occur on the scale required. It is particularly difficult to see how this mechanism could explain the subsidence of closed basins. However, the mechanism may have some relevance to crustal thinning at the time of the initial split between continents. My creep hypothesis (Bott 1971) involves a relatively small extension of the brittle upper crust (about 0-10 km depth range) by normal faulting accompanied by a much greater extension of the ductile lower crust (about 1&35 km depth) by creep. The lower crust is squeezed out from beneath the upper crust at the margin, causing crustal thinning on a much larger scale than is superficially indicated by normal faulting affecting the overlying brittle layer. This type of process can only occur where there is a sink for the outflowing crustal material. One mainstay of my hypothesis is that just such a sink is provided by the sub-oceanic upper mantle beneath the slope and rise at a continental margin, and furthermore stress differences associated with a margin are of the type required to drive such a process, with progressive loss of gravitational energy. The test applied by Matthews & Smith fails to distinguish between the creep