GeoLines
15 (2003)
72
GeoLines
15 (2003)
73
bedded tephra beds of the tephra ring represent the proximal
tephra deposits and point to the number of eruptions of the maar-
diatreme volcano. From the foot of the tephra ring, very thinly
bedded tephra extend outwards in a thin veneer for up to a few
hundreds of km and represent the distal tephra deposits.
Size of diatremes. Geophysical exploration and drilling
as well as investigation of the country-rock clasts in the maar
tephra give evidence that maars are underlain by diatremes (e.g.,
Hawthorne 1975, Lorenz and Büchel 1980, Lorenz 1982a, Bü-
chel 1984, 1987, 1988, Büchel et al. 1987). Exposed diatremes
may be several tens of metres to over 1.5 km wide and less than
100 m to over of 2.5 km deep. In hard rocks as is known from
diamondiferous kimberlite mines in South Africa they frequently
dip inward at average angles of 82° (Hawthorne 1975). The dia-
treme fill consists of volcaniclastics, subsided blocks of country
rocks and a variable amount of intrusive rocks. The volcaniclas-
tics themselves comprise:
1. phreatomagmatic tephra beds in the upper diatreme levels,
especially in the larger diatremes, with characteristics as oc-
curring in the tephra ring, and
2. reworked pyroclastics in the upper but more prominently in
the lower diatreme levels, derived from the tephra ring and
the walls of the maar crater.
3. In addition, the diatreme fill contains phreatomagmatic te-
phra occupying various vertically orientated channels which
represent enlarged original feeder vents through which the
tephra clouds were rising to the surface.
4. The diatreme may also contain large blocks and rock slices of
country rocks subsided from higher stratigraphic, resp. high-
er structural levels compared to the country rocks in the dia-
treme walls.
Historic maars. Despite the frequency of maar-diatreme volca-
noes in volcanic fields and other volcanic environments, only
very few maar-diatremes formed in historic times. The most
recent ones are the Nilahue maar, resp. Carran, which formed
in Chile in 1955 (Müller and Weyl 1956, Illies 1959), Iwo Jima,
Japan, formed in 1957 (Corwin and Foster 1959), the Ukinrek
Maars, which formed in Alaska in 1977 (Kienle et al. 1980,
Self et al. 1980, Büchel and Lorenz 1993, Ort et al. 2000)
and the Westdahl maar which formed on the Aleutian Islands
in 1978 (Wood and Kienle 1990). The maars best studied du-
ring and after their eruptions were the Ukinrek Maars (Kienle
et al. 1980, Self et al. 1980, Büchel and Lorenz 1993, Ort et
al. 2000). The Ukinrek West Maar formed within 3 days; its
crater was finally 175 m wide and 35 m deep. Ukinrek East
Maar formed in the following 8 days and, after its syneruptive
growth, reached a diameter of 340 m and a depth of 70 m. Both
Fig. 1. Schematic diagram of a maar-diatreme volcano show-
ing its feeder dyke, root zone, overlying cone-shaped
diatreme (with the lower level showing unbedded volca-
niclastics and the upper level showing a saucer-shaped
structure with primary pyroclastic beds interbedded
with beds derived from reworking of mostly tephra-
ring pyroclastic beds), an unconformity in the bedded
sequence because of a collapse phase, feeder vents in
the centre of the diatreme, the maar crater with its pos-
teruptive background sediments and debris flow and
turbidite beds, as well as the proximal tephra ring and
the distal tephra veneer. Scale: width equals depth.
Fig. 2. Schematic growth of a maar-diatreme volcano. Du-
ring phreatomagmatic explosions the associated shock
waves fragment the country rocks in what becomes
an explosion chamber and, less intensively, the sur-
rounding country rocks. A number of such explosion
chambers form the irregular-shaped root zone. The dia-
treme and the maar crater grow by downward pene-
tration of the root zone and consequent phases of
collapse, resulting in a larger and deeper diatreme and
a larger and deeper maar crater. Scale: width equals
depth.
GeoLines
15 (2003)
74
GeoLines
15 (2003)
75
craters were filled more (East Maar) or less (West Maar) with
a lake. Judging from the comparison of the life span of active
scoria cones and maars and their respective sizes, it is highly
conceivable that maars with large diameters and high depths of
the respective maar crater were active for months or even years
and that maars grow in size the longer they are active (Lorenz
1985, 1986).
Formation of maar-diatreme volcanoes
The phreatomagmatic model. The complex process chain in
the formation of maar-diatreme volcanoes has been analysed in
a number of recent papers (Lorenz 1985, 1986, 1998, 2000a–c,
Zimanowski 1986, 1992, 1998, Lorenz et al.1990, 1994, 1999,
2002, Fröhlich et al. 1993, Zimanowski et al. 1995, 1997a–c,
Büttner and Zimanowski 1998, Ort et al. 2000, Lorenz and
Kurszlaukis in press) and thus will be reviewed only briefly
here.
In principal, the maar-diatreme volcanoes form when
magma, irrespective of its chemistry (Lorenz et al. 1994),
rises through the feeder dyke and, close to surface, interacts
explosively with groundwater. The interacting magma–ground-
water volume leads to brittle fragmentation of the involved
magma volume and the consequent shock waves generated
by these thermohydraulic explosions have the quality to frag-
ment the surrounding country rocks (Zimanowski et al. 1997c,
Kurszlaukis et al. 1998, Zimanowski 1998, Lorenz et al. 1999,
2000, 2002, Lorenz and Zimanowski 2000). The generation
of water vapour from the interacting groundwater leads to
further fragmentation of the magma surrounding the interact-
ing magma–groundwater volume and to the rise towards the
surface of the eruption cloud thus generated. At the surface,
further decompression of the eruption cloud towards ambient
pressure and consequent condensation of large amounts of wa-
ter vapour lead to fallout of the dominant part of the tephra and
its deposition on the crater floor and on the surrounding surface
by base surges, ballistic transport and some minor tephra fall.
Depending on the pre-eruptive topography and interstitial water
derived from the phreatomagmatic eruptions and/or from rain-
fall, the tephra-ring deposits may collapse during crater growth
and form lahars, thus thicker accumulations on the crater floor.
If deposited on steep relief outside the crater, these deposits
may also form lahars in nearby valleys or in other depressions.
The distal tephra deposits in the near vicinity of the tephra ring
may still contain base-surge material and some ballistics but
going outwards, depending on the wind activity, ash falls (and
rapidly decreasing lapilli falls) from eruption clouds drifting
away from the maar crater rapidly dominate the distal deposits.
These distal deposits may extend up to several hundreds of km
away from the maar crater.
The primary thinly bedded base surge, ballistic and tephra-
fall deposits inside the diatreme are interbedded with thick te-
phra beds that represent redeposited tephra, i.e. sediments, vol-
canogenic debris flows and mudflows. These lahars are derived
from the moist thinly bedded pyroclastic beds by syneruptive
collapse of arcuate slices of the tephra ring onto the inner crater
walls and consequent flow onto the floor of this depocentre.
The actual thermohydraulic explosions occur at the top end
of the feeder dyke and result in a near-spherical fragmented
space, the so-called explosion chamber (Fig. 1 and 2; Lorenz
2000a, c, Lorenz and Zimanowski 2000, Lorenz and Kurszlau-
kis in press). Partial evacuation of the clasts and fragmented
dyke magma from the region of the explosion chamber by the
rising eruption cloud leads to a mass deficiency above the feeder
dyke. Downward penetration of the explosion site (Lorenz 1985,
1986, 1998) leads to a series of interconnected explosion cham-
bers one below another and in part also laterally next to each
other (following mostly the trend of the feeder dyke) and thus to
the irregularly shaped root zone (Fig. 2). The total mass deficit
in this root zone ultimately leads to increasing rock mechani-
cal instability of the overlying country rocks and the diatreme.
After this instability exceeds a critical value, a collapse of
the overlying rock volume into the root zone takes place in order
to compensate the mass deficit. The collapsing material – in
response to the subsiding volume of rocks overlying the root
zone – represents a cone of subsidence like a sinkhole; i.e., the
diatreme previously overlying the root zone top end, with its
lowermost part reaching into the root zone, collapses further
down into the root zone and by this process engulfs the top part
of the root zone (Lorenz and Kurszlaukis in press). The root zone
consequently looses its former upper levels. Since the zone of
subsidence propagates upward towards the Earth’s surface, the
Earth’s surface also has to subside and a crater of subsidence
origin is consequently formed: the maar crater (Fig. 2). Renewed
explosions, further downward penetration of the root zone and
more eruptions result in a subsequent phase of increasing mass
deficit and collapse. As long as magma–groundwater interac-
tion continues, these repeated collapse phases result in repeated
growth phases of the conically shaped diatreme and also in re-
peated growth of the maar crater (Fig. 2).
From the first collapse phase onwards the maar crater
represents a depocentre collecting crater-wall debris falling
and flowing onto the crater floor. As long as the maar crater is
rather small, the growth of the crater results in undercutting of
crater-walls and of the inner parts of the overlying tephra ring.
Collapsing country rocks, i.e. rock falls, rock slides, scree, and
lahars from collapsing arcuate slices of the tephra ring will col-
lect on the crater floor. From a certain size of the maar crater
onwards the crater floor will also represent a depocentre for
pyroclastic deposits, i.e., base surge and impact as well as a few
tephra fall deposits. Thus primary tephra beds, country rock de-
bris of various kind, and reworked tephra will form alternating
beds on the crater floor and jointly subside towards depth in
the growing diatreme and consequently they will get overlain
by further primary pyroclastic deposits and sediments derived
by collapse of the tephra-ring deposits and underlying country
rocks. Note that the site where bedded tephra occur now in a
diatreme is the site reached after subsidence and not the site
where they had been deposited originally.
The magmatic model. It has to be stated, however, that – in
contrast to this author’s model – a number of authors favour
a magmatic model especially for the formation of kimberlite
and carbonatite maars and diatremes and relate their formation
to very volatile-rich magmas and aspects of explosive exsolution