Jökull - 01.12.1985, Síða 6

stratospheric plume 650 km downwind from the volcano
(.Inn et al. 1982) where HCl was estimated at 10—50
ppbv and methyl chloride (CH3 Cl) up to 5 ppbv, and a
HC1/S02 ratio of 0.1 to 0.5. Chlorine is widely reported,
both in fumarolic gases issuing from the crater floor
(Evans et al. 1981) and air-borne filter samples over
2000 km from the volcano (Sedlacek et al. 1982).
Finally, the petrologic chlorine estimate falls in the mid-
range of the estimate of Turco et al. (1982) of 1010 to
10n g C1 total emission during the eruption. Data from
Mount St. Helens (Inn et al. 1982) and some other small
eruptions (Lazrus et al. 1982) indicate, however, that
most chlorine is fractionated from the plume before
chlorine-bearing compounds stabilize as volcanic aero-
sols in the stratosphere. Whether such fractionation
occurs in chlorine-rich large explosive eruptions such as
Tambora 1815, with a mass eruption rate three orders of
magnitude higher than Mount St. Helens, is not known
at this time. The fate of the chlorine in the eruption
column is highly dependent on its molecular form. Thus
the highly soluble HCl may combine with atmospheric
water vapor in the troposphere and be rapidly removed
by precipitation. Methyl chloride (CH3C1) is, on the
other hand, relatively inert in the troposphere and may
therefore be transported to the stratosphere where it is
dissociated to chlorine radicals by ultraviolet radiation.
The results in Table 1 indicate a significant degassing
of fluorine during the Mount St. Helens eruption. The
fluorine data should be treated with great caution,
however, because of the poor detection limit for this
element with the microprobe (300 ppm). HF was
observed in considerable quantities in fumarole gases
emanating from the crater floor (Evans et al. 1981) and
fluorine was also present in significant amounts as acid
adsorbed on particles in the tephra deposit (Nehring
and Johnston 1981) as well as in the down-wind plume
(Hobbs et al. 1981). Neither CFC13 nor CF2C12, how-
ever, were found to be above normal stratospheric
concentration in the distal plume (Inn et al. 1981). By
consideration of these observations, we judge that our
fluorine yield from this eruption is either an overesti-
mate, or that nearly all fluorine was rapidly deposited
by adsorption on tephra particles. The latter process
was well documented during the 1970 explosive erup-
tion of Hekla volcano, where 3xl010 g fluorine were
precipitated near-field with the tephra (Oskarsson
1980).
SOUFRIERE 1979
A petrologic estimate of the volcanic volatile yield
from the eight explosive eruptions of Soufriere volcano
(St. Vincent) in 1979 can be made on basis of the data of
Devine et al. (1984). The magma had a complex crys-
tallization history before eruption, resulting in the trap-
ping of a range of liquid compositions in olivine and
plagioclase phenocrysts (Devine and Sigurdsson 1983).
Late liquids trapped by plagioclase are more representa-
tive of the magma which degassed to give rise to the
volcanic aerosol and hence are used here in mass esti-
mates (anal. 21 and 22 in (Devine et al. 1984)). Scaling
to the total erupted mass of 4x 1013 g of juvenile tephra
(Sigurdsson 1982b), we calculate a sulfuric acid yield of
1.5xl010 g and HCl yield of 3.5xl010 g to the atmo-
sphere, or total yield of 5xl010 g during the eight
explosive events. By comparison, the volcanic aerosol
mass determined by SAGE satellite measurements for
only two of the eruption plumes is 2.3xlO9 g (McCor-
mick et al. 1981). It is likely that the satellite-based
aerosol extinction measurements include about half of
the erupted mass, in which case the petrologic estimate
of volatile degassing is higher by an order of magnitude.
Both sulfuric acid droplets and chlorine-bearing parti-
cles were collected at the periphery of one of the
eruption clouds (Woods and Chuan 1982), and high
chlorine content reported in both filter samples and
tephra (Sedlacek et al. 1982).
The petrologic estimate of sulfur yield during the
explosive phase of the eruption indicates a yield of 120
ppm sulfur per mass of erupted magma (Devine et al.
1984). Direct measurements of sulfur mass flux emis-
sion with COSPEC (Hoff and Gallant 1980) during the
subsequent effusive or lava dome growth phase of the
eruption (May to October 1979) indicate nearly the
same yield. Following the last explosive event on 26
April 1979, volcanic activity continued as magma was
extruded, initially at a rate of 1.25x 1012 g/day, to form a
lava dome (Huppert et al. 1982). during this period, S02
was vented to the troposphere at a rate of 3.4x 108 g/day
(Hoff and Gallant 1980). This sulfur yield of 136 ppm
per erupted mass during the effusive phase is therefore
closely comparable to the petrologic estimate of 120
ppm sulfur yield during the explosive phase, lending
further credence to the petrologic method of the study
of volatile release.
SANTORINI 1470 B.C.
In about 1470 B.C. the Mediterranean volcano San-
torini (Thera) produced 6.7X1016 g tephra (Watkins et
al. 1977, Sparks and Huang 1980, Wilson 1980). We
have estimated a volcanic volatile yield of 3.86X1012 g
H2S04 from this explosive eruption, on the basis of glass
inclusion petrology. No loss of chlorine is indicated
from these data, although the chlorine content of the
magma was quite high (Devine et al. 1984). Hammer et
4 JÖKULL 35. ÁR