J. Barclay et al., PRE-ERUPTIVE VOLATILE CONTENT AND DEGASSING HISTORY OF AN EVOLVING PERALKALINE VOLCANO, Journal of volcanology and geothermal research, 74(1-2), 1996, pp. 75-87
The volatile contents of melt inclusions trapped within phenocrysts of
quartz and feldspar in peralkaline rhyolites from Mayor Island, New Z
ealand, have been studied using Fourier transform infrared spectroscop
y and ion microprobe analyses. The glass inclusions analyzed span the
similar to 130 ka subaerial eruptive history of the island and come fr
om volcanic deposits representing a wide range of eruptive styles (exp
losive vs. effusive, magmatic vs. phreatomagmatic, low vs. high discha
rge rates). The water content of all inclusions analysed is uniformly
high at similar to 4.4 wt.% H2O, whereas CO2 contents are below the li
mits of detection (similar to 50 ppm). Chlorine in melt inclusions ran
ges from 2070 to 5200 ppm, while coexisting matrix glasses have genera
lly lower Cl concentrations of 1700-4200 ppm; the apparent bulk distri
bution coefficient describing Cl partitioning between melt and vapor p
hase ([Cl](fluid)/[Cl](melt)) during degassing is 5-15. Fluorine appea
rs to be less affected by eruptive degassing than is Cl, and melt incl
usion and matrix glass F concentrations show significant overlap (1400
-2520 ppm F in inclusions, 1550-2890 ppm F in matrix glass). The obser
ved invariance of melt inclusion water content with sample age contras
ts with ion microprobe data on incompatible trace elements (e.g., Zr,
Nb), which suggest similar to 35% fractionation between the oldest and
youngest samples. This, along with the Cl partitioning behaviour, sug
gests that at least the upper erupting portion of the Mayor Island mag
ma chamber was water saturated throughout the volcano's 130 ka eruptiv
e history. Furthermore, the large range of eruptive styles observed on
Mayor Island are not due to differences in water concentration of the
erupting magma but instead must reflect differences in rates of magma
ascent and supply (which control the efficiency of non-explosive dega
ssing).