The Wilis Volcanic Complex ONVC) and Lawu Volcanic Complex (L VC) are two volcanoes situated in the eastern part of the island of Java, Indonesia. The volcanoes are separated by an east-west distance of about 70 km, and are part of the chain of volcanoes which forms the eastern Sunda volcanic arc, a subductionrelated feature that developed along the convergent boundary between the IndianAustralian oceanic plate to the south and Eurasian continental plate to the north. The eruptive products of the WVC are classified into five groups based on morphostratigraphy and lithology, these are, from oldest to youngest, the precaldera Klotok basalt and basaltic andesite, the pre-caldera Pawonsewu andesite, the caldera Ngebel high-Si andesite and dacite, the post-caldera Jeding basaltandesite, and the post-caldera Argokalangan andesite. Estimated temperatures of crystallisation of WVC magmas are about 1150 0C for the basalts, about 1060 0C for one olivine-bearing andesite, and range from 1035 0C to 950 0C for the andesites. Water contents of the magmas seem to have increased from basaltic through andesitic to dacitic. The bulk compositions of the WVC andesites suggest that the magmas may have formed in a relatively deep magma chamber, at depths of 7 - 18 km (equivalent to pressures of 2 - 5 kb), with small amounts of water. The WVC basalts may have began to cry stallise at even greater depths, presumably around 30 km (equivalent to pressures,.. 8 kb), and the dacites may have fanned at about 18 km depth, possibly at slightly greater depths than the most andesites. The geochemical and isotopic characteristics of both the pre-caldera Pawonsewu and post-caldera Argokalangan andesites are consistent with an origin largely by crystal-liquid fractionation from the pre-caldera Klotok basalt parental magma. Early fractionation of Ol+Cpx+Pl+Mg assemblages followed by Opx+Cpx+Pl+Mg±Apt assemblages could have produced the pre-caldera Pawonsewu and post-caldera Argokalangan andesites from Klotok-type basaltic parental magma. However, whereas both the pre-caldera Pawonsewu and postcaldera Argokalangan andesites seem to have formed under open-system conditions, with the magma chamber being replenished by new magmas, assimilation of some crustal material may have been involved during the generation of the pre-caldera Pawonsewu andesite. The origin of the caldera N gebel high-Si andesite and dacite is more difficult to interpret. Major, trace and REE characteristics of the dacite (including High-Si andesite) are not consistent with an origin by simple fractionation from andesite. They may be related petrogenetically to the pre-caldera Klotok basalt/basaltic andesite, despite a silica-gap of about 7 wt % Si02. Crystal fractionation involving the separation from Klotok-type basaltic parental magma of assemblages of O l+Cpx+Pl+Mg followed by either Opx+Cpx+Pl+Mg, Pl+Amp+Mg or Opx+Pl+Amp+Mg, could explain the major element variation from the basalt to dacite. However, 87 Srt86sr and o18o values increase with increasing Si02, Rb and Rb/Sr, and decrease with increasing MgO and Sr contents, suggestive of some crustal contamination during differentiation, and trace element modelling and the extremely high o18o (12.1%o- 13.5%o) values of the amphibole phenocrysts from the dacite suggest that even more complex differentiation processes may have taken place. The L VC is divided into two groups, i.e. Old La wu which is characterised by hornblende-bearing andesites and Young La wu which is composed of olivinebearing andesites. Unlike Wilis, basalts are not found. The estimated temperatures of crystallisation of the LVC andesites range from 125QOC to 9820C for Old Lawu and from 132QOC to 9860C for Young Lawu. Amphibole-bearing andesite may have formed at depths of 6 - 7 km, whereas the more basic andesite could have formed at greater depths, perhaps up to 18 km. Old Lawu andesite seems to have developed by crystal fractionation in an open-system magma chamber with the magma chamber being replenished b y new magmas. Crystal fractionation could also account for the geochemical variation within the Young Lawu andesites, although contamination may have also been partly responsible for the geochemical variations. It is not known whether both Lawu groups came from the same parental magma, or even if their parent was basaltic. A more speculative interpretation suggests that the Young Lawu andesite may have resulted from fractionation of a mantle-derived basaltic parental magma, whereas the Old Lawu andesite more probably derived from a basaltic andesite parental magma produced at the upper mantle - lower crust as a result of reaction between mantlederived magma and anhydrous ferromagnesian phases. Zr/Nb, 8 7sr;8 6sr and 143;144Nd values of the WVC basalts and the 87sr;86sr and 143;144Nd values of LVC andesites suggest that MORB mantle exists in the mantle wedge beneath these volcanoes. The REE patterns of the WVC basalts and Young Lawu andesites are inconsistent with models calling for residual amphibole in the source region, but may be adequately modelled by melting of spinel lherzolite. Wheller et al (1987) identified four regional along-arc sectors in the magmatism of the Sunda-Banda arc. They placed the transition from their West Java to Bali sector between Lawu and Wilis and proposed that the geochemical and isotopic differences reflected a greater involvement of crustal material in the sources of the magmatism. It is here argued that these differences arose because the effects of crustal assimilation are more pronounced in the Lawu Volcanic Complex than in the Wilis Volcanic Complex, possibly because the arc crust is thinner under Wilis than beneath Lawu.
History
Publication status
Unpublished
Rights statement
Copyright the Author - The University is continuing to endeavour to trace the copyright owner(s) and in the meantime this item has been reproduced here in good faith. We would be pleased to hear from the copyright owner(s).