S339 Understanding the Continents

Block 2

From Rifting to Drifting

SUMMARIES


The origin and evolution of basaltic magmas

Basalts - classification

A basalt is defined as a dark, fine-grained mafic igneous rock with a SiO2 content between 45 and 52 wt % and less than 5% Na2O + K2O.

Modal analysis of erupted rocks such as basalts has limited use due to wide variations in crystal content resulting from local cooling conditions rathre than being a fundamental property of the magma.

An alkali-silica plot (Na2) + K2O versus SiO2) effectively distinguishes alkali basalts from subalkaline basalts.

Normative analysis takes the major element composition of a basalt and recasts this into a set of standard anhydrous minerals.

Normative analysis classifies basalts by their degree of silica saturation: alkali basalts (undersaturated and Ne-normative), olivine tholeiites (saturated and Hy-normative), and quartz tholeiites (oversaturated and Q + Hy-normative).

Basalts which plot as alkali basalts on an alkali-silica plot will usually be classified as alkali basatls using their narmative mineralogy. The same applies to tholeiitic basalts, although normative classification allows their further subdivision into quartz and olivine tholeiites.

Basalts - petrogenesis


Both alkaline and subalkaline basalts erupt in the ocean basins. Ocean ridges are almost exclusively characterized by tholeiites whereas both tholeiites and alkali basalts erupt on large ocean islands. Small ocean islands are dominated by alkali basalts.

Phase diagrams illustrate the importance of eutectics and thermal divides in controlling the evolution of residual liquid compositions produced during cooling of basalt melt.

A thermal divide (represented by the CPSU in the normative basalt tetrahedron) acts as a ridge or barrier at low pressures (less then 0.5 Gpa) between residual liquids formed during franctional crystallization of alkali basalts and tholeiites.

Mantle Melting


Although the mantle undergoes solid-state convection, variations in peridotite composition suggest that it is not cehmically homogeeous.
Lherzolite is the dominatn upper mantle composition, and when it partially melts it produces a liquid (ie basalt melt) and leaves a solid residue depleted in clinopyroxene.

The upper mantle is mineralogically layered, and with increasing depth (pressure) plagioclase lherzolite, then spinel lherzolite, then garnet lherzolite are encountered. This change is brought about by the pressure-dependent stability ranges of the Al-bearing phases (plagioclase, spinel and garnet).

The mantle partially melts when it moves from teh sub-solidus field and crosses the solidus into the solid + melt field that lies between the solidus and liquidus curves.

Upwelling of mantle under adiabatic conditions results in decompression melting, two important examples of which are mantle plumes and lithospheric thinning. Mantle plumes are independent of tectonic influence, whereas lithsopheric thinning is intimately linked to global tectonics.

The potential temperatures of the mantle is the temperature it would have if moved to the surface along an adiabatic temperature gradient.

Assuming that melt thickness equals ocean crust thickness allows relationships between potential temperature, the amount of melt generated, and the depth of melting to be established.

The small potential temperature difference (c. 80 deg C) between typical ocean crust that ranges from 5km to 10km in thickness suggests that the upper mantle is kept well stirred by convection.

Basalts - origins


The Mg# of a basalt is analogous to the Fo value of an olivine (or its liquid), and provides a first-order means of assessing whether or not a basalt could represent a primary melt.

Mantle olivines have a restricted range of composition (Fo88-92) and the range of liquids (ie primary melts) in equilibrium with them is also restricted (Mg# = 65-73).

High-pressure experimental work on spinel lherzolite KLB-1 indicates the effect that pressure and temperature have on the compositions of primary melts.

SiO2 is strongly controlled by pressure, whereas MgO is strongly controlled by temperature.

Alkali and tholeiitic basalts have their own primary melts.

Trace elements in basalts


The distribution of trace elements between solid and melt is described by partition coefficients.

Compatible elements have partition coefficients less than 1; incompatible elements have partition coefficients more than 1.

High incompatible trace element concentrations are favoured by small melt fractions, low D-values and high source concentrations.

Alkali basalts are the product of smaller melt fractions than are tholeiites.

The source region of MORB is depleted in incompatible trace elements.

Low HREE concentrations in OIB result from the presence of residual garnet during melting.

The source region of OIB has different incompatible element ratios than the source of MORB.

The Kenya Rift

The geography of the East African rift system
African topography is dominated by braod swells (plateaux) and plains.

There is a lack of linear mountain belts on the African continent.

The East African Rift is located on two plateaux, the Ethiopian Plateau and East African Plateau.

The Kenya Rift cuts through the Kenya Dome, an area of unusually high elevation superimposed on the East African Plateau.

The geology of the Kenya Rift


The East African Rift follows the mobile belts and seldom cuts across the craton, hence the existence of two rifts, Kenya and Western, across the East African Plateau, and only one across the Ethiopian Plateau.

The geology of Kenya comprises three major units, the Precambrian Basement, the Mesozoic sedimentary cover and the Tertiary volcanic rocks.

The basement can be divided into the Archean Tanzanian craton, the Proterozoic mobile belt and the remobilized craton margin lying between the two.

The distribution of Mesozoic sedimentary rocks reflects the location of old rift basins generated during the Mesozoic, possibly co-incident with the drift of Madagascar away from the the east coast of Africa.

The Tertiary volcanic rocks of the Kenya Rift vary widely in composition but are generally mafic (basalts basanites and nephelinites) or evolved (rhyolites, trachytes and phonolites). They show a bimodal distribution of compositions.

Magmatism in Kenya migrates from the north to the south and from the wst to the east through time.

Nephelinites and carbonatites are largely restricted to areas of cratonic basement whereas basalts predominate on the mobile belt.

Throughout much of its length, the rift starts as a broad extensional feature and narrows through time.

Geophysical studies of the Kenya Rift


We can put all of these geohpysical results together to fome a composite picture of the structure of the Kenya Rift (fig 3.19, page 58) and the important points to note are listed below:

The whole crust section has been thinned beneath the rift axis.

The amount of extention decreases from Β = 1.6 in the north (Turkana) to Β = 1.1 in southern Kenya.

The mantle is hot and has slow seismic velocity beneath the rift axis.

The mantl is cold and it has fast seismic velocities beneath the rift flanks.

The boundary between the cold and hot mantle is very steep and extends to the base of the lithosphere.

The lithosphere is >160 km thick beneath the craton and ~125km thick beneath the mobile belt.

The mantle beneath the rift axis contains partial melt. The melt fraction is between 3% and 8%.

Extensional tectonics of the African Rift


Rifts form in response to extensional stress within the lithosphere.

Rifts can either have a full or half-graben structure.

The half-graben is the most common structure in extended terrain.

The Kenya Rift has evolved through a half-graben stage into a full graben.

Alternating half-graben are linked by complex accommodation zones which can act as pathways for magmatism.

Lithosphere extenstion can occur by either pure shear or simple shear.

A model involveing extension by simple shear in the upper crust and pure shear in the loer crust and mantle lithosphere best accounts for the evolution of the Kenya Rift.

Petrogenesis of Kenyan volcanic rocks


The volcanic rocks of the Kenya Rift show a bimodal distribution with large volumes of mafic rocks, large volumes of felsic rocks but few intermediate compositions.

Mafic rocks are generally alkaline and silica undersaturated (Ne-normative), and range in compositions from alkali basalts, basanites and nephelinites.

Alkali basalts, basanites and nephelinites are not related to each other by fractional crystallization but represent distinct mantle-derived magmas.

Alkali basalts, basanites and nephelinites are derived from the mantle as a result of variable degrees of melting at different depths; nephelinites from the deepest and smalles melt fractions, basanites from intermediate depths and melt fraction and alkali basalts from th shallowest depths and larges melt fractions.

Variation in composition within the different mafic rock suites is cominated by olivine fractionation.

Felsic rocks range in composition from rhyolites, trachytes to phonolites.

The felsic rocks of the Kenya Rift can be derived both by partial melting of a suitable source and by fractional crystallization of a suitable parental basalt.

The geodynamics of the African Rift


The African Plate is surrounded by constructive plate boundaries which exert a modest compressional force (ridge-push).

Extension across teh East African Rift must therefore be driven by forces generated within the African Plate.

The African continent is surrounded by old, cold and dense oceanic lithosphere which exerts an extensional force on the ocean ridges and the continental lithosphere.

The continent is in a general state of tensional stress.

This extensional stress is large enough to overcome the strength of the lithosphere only in the most elevated parts of the continent.

These elevated areas are also magmatically active and have a high geothermal gradient so that the lithosphere is weaker and thus fails more easily under tension.

The mantle plume has a two-fold effect in that it provides dynamic elevation and thermally weakens the continental lithosphere.

The Red Sea

The regional geology and tectonics of the Red Sea

The Red Sea lies in an elongate basin within late Proterozoic continental lithosphere which developed as a rift during the Tertiary.

If Arabia is rotated clockwise from its present position relative to Africa, the Red Sea can be closed, but there are large areas of overlap both in the north and in the south.

At the southern end of the Red Sea, a triple junction is formed from the intersection of the Red Sea Rift, the Gulf of Aden and the East African Rift.

The southern Red Sea basin lies in a region of domal uplift and has widespread post-30 Ma basaltic magmatism exposed on both its uplifted flanks.

The Red Sea basin is floored by 4 km of evaporites which provide evidence of subsidence in an enclosed basin between 20 Ma and 5 Ma.

Flanking conglomerates in the Red Sea basin provide evidence for an escarpment developing since 20 Ma.

Basaltic magmatism in the Red Sea

Basaltic magmatism in the Red Sea Region started at ~30 Ma with the eruption of the Ethiopian and Yemeni flood basalts.

The Ethiopian flood basalts are of transitional composition and erupted over a short period of time (1.6-2 Ma).

The high -Ti (HT2) magma types were sourced in the Afar plume whereas the low-Ti (LT) magmas were derived from shallower levels, probably in the mantle lithosphere.

The Afar Depression is a region of active extension, floored by transitional and tholeiitic basalts that date from 25-30 Ma to the present day.

Basaltic magmatism on the Arabian peninsula comprises flows (harrats) and dykes that range in age from 27-29 Ma and 15-0 Ma.

The harrats and oldest dykes are alkaline whereas younger, coast-parallel dykes are tholeiitic.

Basalts from the Gulf of Aden and the Red Sea axial trough are low-K tholeiites (MORB).

Continental or oceanic lithosphere?

The southern Red Sea comprises a broad basin with a deep axial trough.

The axial trough is floored by basalts and has large magnetic anomalies; it is true oceanic crust.

The rest of the Red Sea is floored by evaporites, beneath which is attenuated continental crust, intruded by mafic dykes.

The Afar Depression is floored by basaltic rocks with a composition distinct from those of the Red Sea axial trough.

The crust beneath the Afar Depression is attenuated continental crust, intruded and thickened by basaltic intrusion.

The northern Red Sea does not possess a well-defined axial trough and is attenuated continental crust throughout, again with some dyke intrusion.

The Gulf of Suez is a half-graben in continental crust; amounts of extension are low.

The timing of extension in the Red Sea and Gulf of Aden

Fault displacement of the laterite horizon in Eritrea reveals subsidence of the Red Sea basin and flank uplift caused by flexural rebound during the early Neogene.

Sedimentary evidence indicates formation of the Red Sea Rift between 29 and 15 Ma.

Fission track evidence points to two episodes of rifting, uplift and denudation, at 30-35 Ma and 20-25 Ma.

Both events are recorded in both the northern and southern Red Sea although only the older is recognized in the Gulf of Aden.

The data suggest simultaneous rifting along the length of the Red Sea.

Both Arabia and Africa acted as rigid blocks during rifting.

Plate kinematics

The Red Sea region can be divided into three major plates: the Arabian, African and Somali Plates.

These are separated by constructive margins of the African Rift, with an RRR junction located in the Afar Depression.

Differential extension across the Gulf of Suez and the Red Sea basin is accommodated by the Dead Sea transform.

The Arabian Plate is moving counter-clockwise (north-east) about a pole of rotation in the southern Mediterranean.

Hot spot trails in the southern Atlantic show that the united Afro-Arabian Plate was moving counter-clockwise (NE) at about 30 mm yr-1 from 80 to 30 Ma.

The African Plate slowed significantly between 30 and 20 Ma, at the same time as the Red Sea first rifted.

This slowing can be related to the isolation of the African Plate from the slab-pull of subduction beneath the Zagros mountains.

Rifting of the Red Sea may have been partly triggered by the initial emplacement of the Afar mantle plume

Flood basalts and continental break-up

Flood basalts show a close spatial and temporal association with many phases of the break-up of Gondwana.

As break-up proceeds, the time gap between flood basalt magmatism and sea-floor spreading decreases.

Different flood basalt provinces have different characteristics, especially their duration, eruption rate and mantle source regions.

Early flood basalts such as the Parana are dominated by lithospheric mantle sources, whereas later provinces (eg the Deccan) show more involvement of mantle plume sources.

The lack of continental break-up associated with one of the largest flood basalt provinces (Siberia) implies that mantle plumes cannot alone lead to break-up and sea-floor spreading.

The lack of evidence for mantl plume activity in the rifting of Australia from Antarctica implies that mantle plumes are not essential for continental break-up.

Back to S339