Recent parameterisations of melting experiments may be used to predict the volume and composition of magma generated by melting anhydrous peridotite during adiabatic decompression of the asthenosphere (McKenzie & Bickle 1988). Here they are used to predict the character of magmas generated during adiabatic upwclling which accompanies the formation of sedimentary basins. The objective of this thesis is to test the model predictions with observations from the Mesozoic North Sea Basin. At ocean ridges the degree to which the asthenosphere melts is principally governed by its potential temperature (Tp) and the entropy change on melting (AS). The amount of melt produced and its average composition are controlled by the average extent and average depth of melting. The model predicts that in order to generate the normal 7 km of oceanic crust at a mid-ocean ridge, the 'normal' Tf of the asthenosphere must be 1280±40°C; the error results from uncertainty in AS. Predicted average melt compositions agree favourably with compositions of primitive MORB (mid-ocean ridge basalt). Some ridge segments are located above anomalously hot 'plumes' in the mantle. The greater thicknesses and distinctive compositions of the crust in such regions (e.g. Iceland) are consistent with model predictions for Tp's 100 to 300°C greater than normal. The generation of melt from dry peridotite during rifting of the litho-sphere depends critically on the T,, of the asthenosphere, the amount (J3e) of lithospheric thinning, and the initial shape of the geotherm; i.e. the thickness of the mechanical boundary layer (MBL) prior to extension. In most extensional basins the steady state MBL is ~100 km thick. The maximum amount of lithospheric thinning which occurs during extension depends on the partitioning of strain between the crust (ftc) and mantle (/3m) and on whether extension is accommodated by pure shear or simple shear. At normal Tj,'s the amount and average composition of the melt produced during rifting is simply related to the size of the average melt fraction and the average pressure of melting. When a plume accompanies rifting, regional doming over an area 1000 to 2000 km in diameter is expected. High temperatures and rates of plume-flow may lead to the production of large volumes of alkaline magma when there is little or no stretching and flood tholeiites when ft is large. The most important episode of extension in the North Sea took place from the Middle Jurassic (180-170 Ma) to the Early Cretaceous (120-100 Ma) and resulted in the development of a trilete rift system. Associated mag-matic activity occurred shortly after the onset of rifting (ca. 160-150 Ma) and was located principally in the triple-junction area, where an average thickness of 0.5-1.5 km of subaerially erupted alkali basalt comprise the Forties volcanic province. More minor volumes of nephelinites, basanites and ultrapotassic rocks, are observed on the western edge of the Egersund Basin, on the flanks of the Central Graben and in the Netherlands sector of the North Sea. Dates generated by 40Ar/39Ar stepwise degassing of the freshest samples suggest that the igneous activity is contemporaneous throughout the region; i.e. there is no migration of magmatism with time. The North Sea rocks are variably altered and range from mildly to highly alkaline in character. The Forties basalts are chemically similar to OIB (ocean island basalt), but the ultrapotassic rocks found elsewhere in the North Sea have no oceanic equivalents. All of the North Sea rocks lie between MORB and bulk earth in terms of U3Nd/U4Nd (0.51265-0.51285) and 87Sr/86Sr (0.703-0.704). Simple single stage melting calculations suggest that if the source of the Forties parental magmas was a garnet peridotite with a composition betwen that of bulk earth and the inferred source of MORB, then they were produced by very small amounts of melting (< 3%). The extreme fractionation of LILE (large ion lithophile elements) and light REE (rare earth elements) from heavy REE seen in the magmas from other parts of the North Sea is difficult to explain by single-stage melting unless the source is more enriched in LILE and light REE than bulk earth. The stretching factor (ft) in the triple junction region is estimated from thermal subsidence and gravity data to be between 2 and 3; larger than elsewhere in the North Sea. Many of the subsidence curves are abnormal in that they show condensed, or non-existent, syn-rift subsidence, and very rapid thermal subsidence. These effects are not considered to be due to fault block rotations and do not affect the estimates of ft. They may be explained by decoupling of the strain rates in the crust and mantle which results in extreme but localised thinning of the mantle at the start of rifting (ftm > > ftc) but importantly, allows the lithospheric section to be balanced at all times. The stratigraphy and the absence of a hot-spot trail during the Mesozoic is inconsistent with a plume-model for rifting. There was, however, localised uplift along the rift flanks and in the triple junction, which can be explained by the 'decoupled' stretching mechanism. The simplest model for the Forties basalts involves stretching by a factor of between 2 and 3 of a MBL 90 km thick over asthenosphere with a Tp of 1280 to 1320°C. The predicted volume and composition of melt are largely consistent with these observations. A full explanation of the Forties magma compositions does, however, require an input from mantle that is enriched in incompatible elements and volatiles relative to bulk earth. This component in the forties magmas was probably produced in a metasomatised layer at the base of the MBL. Volatile-controlled melting in the MBL was probably also responsible for the ultra alkaline rocks found in less-stretched regions of the North Sea.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:653696 |
| Date | January 1990 |
| Creators | Latin, David Michael |
| Publisher | University of Edinburgh |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/1842/12393 |
Page generated in 0.0442 seconds