Search

Anatahan Island is located at the southern end of the Mariana volcanic chain. On May 10, 2003, the eastern crater of the island erupted for the first time in recorded history. The Plinian eruption column reached an altitude as high as 13 km on May 11, and a thick layer of ash covered the island. The eruptive activity continued to June 2003, but most of the erupted material was expelled during the first week. The volcanic activity declined in the second half of 2003, but resumed in April 2004. In order to determine crustal deformation associated with the eruption, we conducted GPS measurements in July 2003 at a benchmark (ANAT) located approximately 7 km west–northwest of the active crater, where GPS campaign measurements had been repeated four times since 1992. In the period from January to July 2003 during the eruption, significant subsidence – as much as 21 cm – was detected, but horizontal movement was negligible. We began taking continuous GPS measurements at the same site in July 2003 to monitor the transient deformation that was probably associated with magma migration. To assess the spatial extent of the deformation more accurately, we established another permanent GPS site (ANA2) at a site approximately 3 km from the active crater in the northeastern part of the island in January 2004. The coordinates of this time series at ANAT probably show a change in trends at the beginning of 2004. Another subsidence of 2.8 cm and a westward motion of 2.1 cm were estimated to have occurred in the period from July to December 2003. This was followed by an uplift of 5.2 cm and movement in an eastward direction of 1.0 cm in the period from January to June 2004. We developed three preliminarily models of inflation/deflation sources for three different time periods. During the period from January to July 2003, a deformation source was located beneath the ANAT site and acted as a deflation source. Considering the gap in the GPS time series and errors in data (especially after July 2003), we expected that the deformation sources were located beneath the western part of Anatahan Island and not below the active crater.

On 10 May 2003, Anatahan volcano (located at 16°21' N 145°40' E on the Mariana arc) entered its first historical eruptive episode, sending ash to > 12 km into the atmosphere. Abundant accretionary lapilli, quenched pumicetextures, and hydrothermal minerals in the earliest eruptive deposits indicate hydromagmatic interaction and active mining of the pre-eruptive hydrothermal system. Whole-rock compositions of the products erupted within the first week are chemically homogenous, with SiO 2 ∼61 %, MgO ∼2.1 %, K 2 O ∼1.4%, Na 2 O ∼4.1 % and Fe 2 O 3 ∼9.1 %. The products are classified as medium-K andesites with tholeiitic affinity. Slightly more silicic matrix glass compositions (up to 63 % SiO 2 in microlite-rich matrices) overlap with whole rock, suggesting limited crystal fractionation with microlite crystallization responsible for the more evolved residual melt. Decreasing corrected LOI values (2.3-1.4 wt.%) upsection are consistent with waning hydrothermal mineral contributions as the eruption progressed. Oxygen fugacity calculations based on the ferric to ferrous iron ratio of bulk samples indicate an oxidized magma with ANNO ∼+1. Two-pyroxene equilibrium thermometry suggests magmatic temperatures of 1050-1100 °C. Matrix glass volatile contents show a degassed residual melt, with 1 and Ca, Mg, Na, and K ratios in the leachates suggest a hydrothermal fluid contribution to elements present as water soluble salts adsorbed onto ash. Sulfur budget calculations based on estimates of pre-emptive magmatic and residual melt S contents, mass of erupted magma, and total SO 2 output fluxes require an additional source of S other than the erupted magma. Multiple lines of evidence, including high SO 2 emissions early in the eruption, the presence of accretionary lapilli and hydrothermal minerals in the early eruptive deposits, quenched pumice textures, and cation and anion ratios and abundances in ash leachates suggest that a S-rich free volatile phase exsolved from a large magma body. Magmatic volatiles were stored as components of the hydrothermal system (pressurized gases, hydrothermal fluids, and/or hydrothermal minerals) to be remobilized early in the eruption to contribute to the total SO 2 output.

Anatahan Volcano, Northern Mariana Islands, began erupting in May-June 2003. A series of subplinian explosive eruptions of andesite magma began at the Eastern Crater in the eastern part of the summit caldera on the evening of 10 May. Brown tephra was sent mainly westward by strong winds. Small-scale pyroclastic surges were discharged eastward outside the caldera in late May. An andesite lava dome that had once filled the inner crater was fragmented by phreatomagmatic explosions in the middle of June. The phreatomagmatic explosions probably occurred due to interaction of the magma head with groundwater around the crater, and abundant very fine ash (gray tephra) was discharged within the caldera and over most of the island. The volume of eruption products of the May-June eruption was estimated to be 1.4 x 10 7 m 3 dense-rock-equivalent. Erupted pumices and lava are aphyric andesite and are variously colored depending on their vesicularity. The SiO 2 contents of erupted materials decreased slightly with time. The fine gray ash is depleted in alkalies, probably due to leaching by acid hydrothermal fluids during explosions. Seismic activity resumed in late March 2004, and small strombolian-like explosions were repeated in May and June 2004. About half of the inner crater was filled with new scoria and lava.

[1] Repeated GPS surveys in the Mariana Islands show that the Mariana block is moving apart from the Philippine Sea plate. The velocities of the islands relative to the Philippine Sea plate range from about 15 mm/yr in the northern islands to about 45 mm/yr near Guam. The data also suggest convergence rates for the Mariana forearc with respect to the Pacific plate of 35–45 mm/yr at 19°N increasing to 55–70 mm/yr at 13.5°N. In addition, the velocity vectors show a slight north-south expansion of the arc. The estimated location of the Euler pole of the Mariana forearc with respect to the Philippine Sea plate is well south of the geographical point where the back-arc basin narrows to zero width.