Jökull - 01.12.1989, Síða 62

surtseyan tephra grains are very uniform in shape
and typically equant.
FORM PARAMETERS, DEFINITIONS AND
ANALYTICAL METHODS
SAMPLING PROGRAMME
Samples for the analysis of form were selected
with four objectives in mind. The first one was to
enable comparison between eruptions in widely
different physical environments (submarine, subgla-
cial, subaerial). The second objective was to cover a
wide chemical compositional range (basalt to rhyol-
ite). Thirdly, the variation within one tephra layer
was tested by analyzing four samples from a vertical
section through one layer (H4), and by analyzing
samples from three geographically separated locali-
ties for a single tephra layer (H!). Finally a dupli-
cate analysis was carried out on one of the samples
to test the reproducibility of the analytical method.
Table I lists the samples used in the present study.
All the samples were obtained from existing collec-
tions of tephra samples and no field observations
were carried out. The exact location of some of the
samples is not known and is not considered relevant
to the present study.
There are several possible sources of error in the
determination of form of tephra produced in a vol-
canic eruption. A single eruption may change in
character with time. A sample from a single phase of
such an eruption may not represent a full picture.
This source of error may be minimized during sam-
pling by carefully mapping the tephra bed and by
observing intemal changes in structure. Secondly,
the grain shape may play a part in the sorting of
tephra as it is carried away from the point of erup-
tion, so that the form may change with distance
within the same tephra layer. If certain eruptive
mechanisms, such as phreatomagmatic explosions,
affect the grain shape, some effect on the transport
distance may also be expected because grains pro-
duced in an explosion are likely to be ejected rela-
tively high up in the atmosphere, facilitating long
distance transport. Finally, fresh tephra forms a
relatively unstable sediment in many cases and may
be easily transported and redeposited. This secon-
dary process may change the original shape of the
tephra, but should normally be accompanied by
some rounding.
No attempt was made in the present study to
evaluate these possible sources of error, except for
the samples from H, which was sampled at three
localities, and H4 which was sampled in a vertical
section. A combination of detailed field work and
the study of tephra morphometry would clearly be
desirable to minimize or eliminate some of the
indoor and outdoor sources of error.
GRAIN SIZE AND FORM
All tephra samples (Table I) were divided into
whole O size classes by sieving and size classes
retained on 0.0 to +3.0 (yielding a size range of
0.125 to 2.0 mm) were measured in all the samples.
The mean diameter of some measured grains may
fall outside the upper and lower sieve class limits
because of their irregular shape. The <Þ scale was
defined by Krumbein (1934) and McManus (1963):
O = -log2(d/d0)
where d is the grain diameter in mm, d0 is the refer-
ence unit (1 mm) of the dimensionless 4> scale.
After sieving, 20 grains from each size class were
selected randomly. Rock fragments, phenocrysts,
and exceptionally coloured grains were rejected.
Each grain was then mounted on an elongate glass
plate with the plane of maximum projection of the
grain oriented parallel to that of the glass plate. This
plane contains the long (L) and intermediate (I) axes
while the short axis (S, measuring the grain thick-
ness) is perpendicular to the glass plane. The length
of the 3 axes was measured under a binocular micro-
scope employing a 90° rotation about the long axis
of the glass plate to view the S axis. A duplicate
analysis was carried out on one of the samples, 5838
from Snæfellsjökull.
The values for the major axes of each size class
were fed into a computer program which calculates
various form indexes as well as their means and sta-
tistical parameters, and plots form diagrams to give
a visual concept of the measurements. The form
60 JÖKULL, No. 39, 1989