Environmental Engineering Reference
In-Depth Information
information has come from detailed
ow-by-
ow
magnetic stratigraphies
of thick
sections of lava
flows, in which sequences of superimposed lava
flows with the
same paleomagnetic direction have been identi
ed. These directional groups
(DGs) appear to have cooled in a time too short to have recorded secular variation.
These DGs correspond to single eruptive events (SEEs), with a volume often
larger than 1000 km
3
. Their emplacement could have lasted less than a decade
(Thordarson and Self,
2003
).
There are now six
cient data are available
to propose a common model of embedded timescales of volcanism: large igneous
province (LIP) volcanism appears to occur in a highly discontinuous way, with
embedded timescales on the order of 1 to a few million years (full trap emplace-
ment), 100 to 10 kyr (volcanic phase) and 100 to 10 years (SEEs) (
Figure 15.4
).
This model draws in particular from the work of Chenet
et al
.(
2008
,
2009
) on the
Deccan, supported by similar observations from the Columbia, Brito-Arctic,
Karoo
flood basalt provinces for which suf
Ferrar, CAMP and Siberian Traps. It is not always easy to distinguish the
time corresponding to impingement of the plume head from early rifting to drifting
stages in the cases when opening of an ocean basin occurred. In most cases,
when suf
-
cient geochronologic data are available, volcanism is concentrated in a
few major phases, each on the order of 1 Myr in duration or less. Finer analysis
(based for instance on identi
18
13
C isotopic anomalies in correla-
tive marine sections) shows that these major phases lasted on the order of 100 kyr
in duration (e.g. Keller
et al
.,
2012
). These phases contain a number of SEEs, some
larger than 10,000 km
3
! Failure to record signi
cation of
δ
Oor
δ
cant geomagnetic secular variation
implies cooling times of these SEEs (flow fields) on the order of a decade to at
most a century, whereas recognition of small secular variation loops or structure
in the path of a reversing
field implies time constants on the order of centuries
to a few millennia at most (Chenet
et al
.,
2008
,
2009
).
The recognition of DGs/SEEs allows one to propose that the total time during
which a continental
flood basalt was erupted (i.e. without intervening quiescence
periods) was on the order of only several thousand years; this is about 1% of
the total duration. Many authors note a lack of sediment layers, weathered crusts
or paleosols between
in that
it allows one to identify DGs at the decadal timescale, but not the time between
two successive
flows. The DG method is unfortunately
“
irregular,
”
flows with distinctly different paleomagnetic directions. The study
of inter-trappean sediments would seem more promising.
Interpretation of these patterns is important for plume dynamics, melt generation
and eruption, and eruptive tectonics could be a key element in evaluating the
“
flood basalt. Our working hypothesis is that
the same numbers
N
, volumes
V
, and CO
2
and SO
2
contents of two separate
continental
killing potential
”
of a continental
flood basalts could lead to vastly different biotic and environmental
