Jökull - 01.12.1999, Page 95

Table 1.
Density (p) (crushed = cr, effective = eff) and porosity (n) of rock formations
found in Surtsey (Jakobsson and Moore, 1982)
TypeofLayer pcr [103 kg/m3] Porosity [%] peft [103 kg/m3]
Lava (basalt) 3.08 22 2.40
Pillow lava breccia* 3.08 31 2.43
Pillow lava (wet) 3.05 31 2.42
Tephra (dry) 2.78 46 1.50
Tephra (wet) 2.78 38 2.10
*Pálssonc(ai. (1984)
peft = (l-n)pcr + pa/w. peff is density after accounting for porosity, n is porosity,
pcr is the grain density of the rock, pa/w is pore density, pa/w is either 0 or
1000 kg nr3, depending on whether the pores are filled with water or air.
air photos and maps from various times during the
eruption were used (e.g. Jakobsson and Moore (1982)).
FORWARD MODELING
In the cross-section chosen for modeling, at least 5
geologic units with different density are to be expect-
ed on the basis of the eruption sequence in 1963-1967
and subsequent modification by erosion and deposi-
tion. These are: i) Tephra above sea level (dry); ii)
tephra below sea level (wet); iii) subaerial lava; iv)
pillow breccia below sea level and v) sediments. In
addition, the existence of a core of pillow lava was
explored, but its density should be similar to that of
the pillow breccia. Density variations in the upper-
most crust under the island are also possible. Howev-
er, due to the limited length of the profile, such varia-
tions cannot be modeled except for the uppermost
kilometer under the center of the island. The most
probable densities and porosities for each unit are
given in Table 1. They are based on data from the
Surtsey drill hole (Oddsson, 1982) and from else-
where in Iceland (Pálsson et al., 1984).
The modeling software Gravmag (Pedley et al.,
1991) was used to create 2 1/2-D models. A cross-sec-
tion is modeled by splitting it into polygons that strike
at right angles to the profile. The strike length of the
polygons can be varied as well as the size, form and
density. In the case of Surtsey the strike lengths of the
bodies are determined by the diameter of the island and
other constraints on the width of individual units.
Table 2.
Density (effective), assumed porosity and strike length used in the models
a) Upe of Layer p [103 kg/m3] Porosity [%] Half strike length [km]
Lava (basalt) 2.40 22 0.6
Pillow lava breccia 2.43 31 0.6
Tephra(dry) 1.50 46 0.4
Tephra (wet) 2.10 38 1.0
b) TypeofLayer p [103kg/m3] Porosity [%] Half strike length [km]
Lava (basalt) 2.40 22 0.6
Pillow lava breccia 2.40 33 0.6
Tephra(dry) 1.40 50 0.4
Tephra (wet) 2.00 44 1.0
Pillow lava 2.43 30 1.0
c) lypeofLayer p [103 kg/m3] Porosity [%] Half strike length [km]
Lava (basalt) 2.40 22 0.6
Pillow lava breccia 2.43 31 0.6
Tephra (dry) 1.50 46 0.4
Tephra (wet) 2.10 38 1.0
Pillow lava 2.43 30 0.2
Crater filling 2.40 22 0.1
d)TypeofLayer p [103 kg/m3] Porosity [%] Half strike length [km]
Lava (basalt) 2.40 22 0.6
Pillow lava breccia 2.43 31 0.6
Tephra (dry) 1.45 48 0.4
Tephra (wet) 2.10 38 1.0
Sediments 2.30 35 0.2
RESULTS
In Fig. 5, four models of the structure of Surtsey are
presented. The models show that the observed gravi-
ty field can be explained by plausible density varia-
tions within the edifice itself. The largest variations
occur above sea level and, being well constrained
(e.g. Jakobsson and Moore, 1982), they must form
the basis for gravity modeling of the island. We there-
fore consider our models better for exploring the
structure below sea level than those of Cameron et al.
(1992). In contrast to Cameron et al. (1992), we see
no indications of anomalous bodies in the uppermost
crust below the old sea floor. This does not rule out
JÖKULL, No. 47, 1999
93