V42C-16


Essential Material of the March, 31, 2000 Eruption of Usu Volcano
   A. Tomiya, I. Miyagi, H. Hoshizumi, Y. Kawanabe, H. Sato 
     and Geological Research Group for the Usu Eruption


*Tomiya, A (tomiya@gsj.go.jp)
*Miyagi, I (imiyagi@gsj.go.jp)
*Hoshizumi, H (hoshizum@gsj.go.jp)
*Kawanabe, Y (yagi@gsj.go.jp)
*Sato H (hsatoh@gsj.go.jp)

(*) Geological Survey of Japan, 1-1-3, Higashi, Tsukuba, Ibaraki,
305-8567, Japan


The March, 31 eruption is the first and the largest explosion in the
2000 eruptive activity of Usu volcano. It is regarded as a
phreato-magmatic eruption, however, its climactic stage may be nearly
a magmatic eruption.

This eruption produced small amount of pumice, "Us-2000pm", and at
least tens of thousand tons of volcanic ash. In the ash, we found two
types of volcanic glass; one is microlite-rich glass "Us-2000g", and
the other is microlite-free clear glass "Us-2000t". About a half of
the ash is Us-2000g, whereas Us-2000t is few percent. Among these
products, we determined which is the essential materials of this
eruption.

[pumice] Us-2000pm is similar to Us-1977 (the pumice of the 1977-78
eruption) in appearance, however, our analyses shows that magnetite
composition of Us-2000pm is different from that of Us-1977. Therefore,
it is not recycled pumice of Us-1977, but essential material of the
eruption.

[ash] Us-2000t is identified as recycled Toya pyroclastic flow
deposits (ca. 100ka) because of similar glass composition. Us-2000g is
undistinguishable with Us-1977 in glass composition and oxygen isotope
ratio. However, magnetite composition of Us-2000g is different from
Us-1977 and the same as Us-2000pm. Therefore, Us-2000g is essential
material as well as Us-2000pm. Accordingly, about a half of the
products of the 31 March eruption is essential.

We also discuss the eruption processes deduced from geological,
petrographical and geochemical data. Deduced from similarilty in bulk
rock composition and petrographical characters between Us-2000pm and
Us-1977, unimodal magnetite composition of Us-2000pm/g (no evidence
for magma mixing), and existence of 'relict' phenocrysts which was
also contained in Us-1977, we propose that the Us-2000 magma is the
residual magma of the 1977 eruption.

We also estimated the fragmentation depth of the magma by means of
water content in the matrix glass of Us-2000g analysed by SIMS. The
measured water content of 2.5 plus-minus 0.5 wt.% is corresponding to
about 1.5-3.5km deep, if we suppose water saturation and lithostatic
pressure at the fragmentation. This is considerably deeper than the
estimated aquifer. Thus, we propose that the Us-2000 magma was highly
vesiculated and possibly fragmented below the aquifer.