Source mechanism of the tsunami of 2004 in the Indian Ocean: Analysis and numerical simulation

The tsunami in the Indian Ocean caused by the earthquake of December 26, 2004, near Sumatra Island had catastrophic consequences in coastal areas of many countries in this region. Notwithstanding extensive investigations of this phenomenon at various laboratories of the world, the focal mechanism of the aftershock remains unclear. The paper analyzes possible seafloor movements in the source area of the earthquake on the basis of the keyboard model of tsunamigenic earthquakes and describes numerical simulation of the generation, propagation, and runup of water surface waves in terms of this model involving vertical displacements of seafloor “keyboard-blocks.” It is shown that generated tsunami waves are essentially dependent on the combination of keyboard-block movements, which results in an irregular distribution of maximum runups along the shoreline. If the oblique nature of the subduction zone associated with the Sumatra-Andaman earthquake of December 26, 2004, is taken into account, the model results fit well the runup values observed at the Thailand shoreline. It is noted that this model of the subduction zone accounts more adequately for the tsunami wave field pattern in both areas of the Indian Ocean and other water areas such as the region of the Kurile-Kamchatka Island Arc and the Sea of Okhotsk.     

Atmospheric circulation over western Siberia in 1976–2004

Cyclonic and anticyclonic activity over western Siberia for 1976–2004 is considered. Trajectories of motion of baric formations of various genesis and their frequency are analyzed.     

Post-collisional volcanism in a sinking slab setting—crustal anatectic origin of pyroxene-andesite magma, Caldear Volcanic Group, Neogene Alborán volcanic province, southeastern Spain

Caldear Volcanic Group (CVG), a stratigraphically well defined, calc-alkaline rock complex within S^a de Gata in the eastern part of the Alpine Betic mountain chain, S Spain, consists of three distinct formations: Hernández pyroxene andesites, Bujo hornblende-bearing pyroxene andesites and Viuda hornblende-bearing pyroxene dacites–rhyolites. The letter rock formation may have developed through crystal fractionation of mainly plagioclase and pyroxenes, however there is no direct relation between two formations. CVG has a domainal structure with a northeastern domain where Hernández formation is overlain by Bujo formation while Viuda formation is absent, and a southwestern domain where Viuda formation forms the only fractionate after Hernández formation. Hernández parent magma is thought generated through crustal anatexis by dehydration melting of a predominantly amphibolitic source rock complex which was formed by metamorphism from c. 500 Ma volcano-sedimentary parent material. The domainal structure of CVG is explained by compositional variation within this protogenetic complex. Single crystal U–Pb ages of c. 500 Ma to 1800 Ma for inherited zircon support the presence of clastic material of Proterozoic derivation within the original volcano-sedimentary complex. Regional study of syn-collisional rock formations (Alpine nappe complexes) indicate that the collisional tectonic stage in the Betic-Rif orogenic belt took place rather early (25–30 Ma?) and was followed by a stage of rapid regional rock uplift, fast cooling (c. 500°C/my) and extensional tectonics in the period 22–17 Ma. This later tectonic stage was set into motion by slab break-off which set the stage for a high temperature regime in the overlying lithosphere, providing the framework for the crustal melting and magma production responsible for the calc-alkaline rocks of Alborán volcanic province. Miocene zircon with ages ranging from c. 17 to 11 Ma indicate a rather protracted magmatic development prior to eruption at c. 11 Ma. Post-collisional character of Caldear Volcanic Group thus seems well established.     

A method of analyzing polymodal distributions in orientation data

A method of cluster analysis for orientation data is described. The method enables vector statistics to be applied to polymodal distributions. Similar results to those found by modal analysis are obtained.