3.4 Carbonates
single process which can explain all the various phenomena observed in the rock record. Therefore several
dolomitization models have been proposed (see, e.g.,
Shukla and Baker 1988; Tucker and Wright 1990; Purser et
al. 1994; Tucker et al. 1994; Rao 1996) which are here
briefly surnmarized in a simplified way.
The various dolomitization models proposed in the
literature include:
- Sea-water contact, sea-water pumping. Normal seawater is oversaturated with respect to dolomite and
has a Mg/Ca ratio of 5.2. Consequently, sea-water is
an inexhaustible source of magnesium. It appears
reasonable to assume that at the contact of calcareous
sediment with sea-water, or pore-water of similar
chemie al composition as sea-water, part of the Ca++
of the solid calcareous phase can be exchanged for
Mg++ in sea-water. Strong wave motion and tides
pump pore-water into or out of the rock body for
very long time periods. Upward-directed terrestrial
heat flow may aid in rnaintaining rock-internal fluid
flow. The intense dolomitization of many modem
and ancient, more or less isolated reef bodies (oceanic platforms, seamounts and atolls), particularly at
their steep flanks, is usually interpreted by this
model. Similarly, the dolomitization of extensive,
thick and weB bedded shelf carbonates also seems to
require the overlying water body as Mg-source.
Downward-prograding sea water-induced dolomitization probably proceeds more or less simultaneously with the vertical upbuilding of the carbonate
sequence (bank-by-bank dolomitization).
- Brine-reflux model (or seepage-reflux model and
sabkha model with evaporative reflux). Several
dolomitization models are based on evaporation of
sea water in lagoons or coastal sabkhas in arid to
semi-arid climates. The resulting hypersaline brines
can attain a salt concentration sufficiently large for
the precipitation of gypsum (Sect. 6.4.1). Because
gypsum extracts Ca++ from the solution, the Mg/Ca
ratio of the brine far exceeds that of normal sea water
(Mg/Ca = 5.2) and therefore has the potential for initiating dolomitization of calcium carbonate in the underlying rocks. Due to its increased density, the lagoonal brine tends to sink into the underlying rocks
and then to flow back toward the sea (Fig. 3.30b). On
its way through the carbonate body, the alkaline brine
can also dissolve skeletal opaline silica which is later
reprecipitated as chert in the mixing zone between
brine and normal pore water. A slight modification of
this model is gypsum precipitation in a coastal
sabkha and brine reflux to the sea. These models are
widely applied for ancient dolomites if they are associated with evaporites.
- Meteoric-marine mixing-zone model. A number of
rock sequences with dolomites does not show any
evidence of syn-depositional or post-depositional
arid to semi-arid climate. The basic idea of the
151
mixing-zone model is the experimentally tested degree of saturation of dolomite and calcite in different
mixtures of sea water and fresh water. Normal sea
water is supersaturated for both calcite and dolomite.
When CO 2 -saturated meteoric groundwaters (95 to
50% of the solution) mix with sea water, the resulting solution becomes undersaturated for calcite, but
is still supersaturated for dolomite. Under these conditions, dolomite may form. This process is possibly
furthered by low sulfate concentrations due to the
dilution of sea water and/or bacterial sulfate reduction. In humid c1irnates, lagoonal water can become
brackish or fresh and, due to its lower density, form a
lens-shaped water body reaching into the subsurface
(Fig. 3.30c). In the mixing zone calcium carbonate is
replaced by dolomite (early dolomitization). Similarly, fresh groundwater from emerged platforms, or
from the continent can flow seaward and, dependent
on the hydraulic head, shift the mixing zone or partially flush out the primary marine pore water. The
mixing zone between meteoric water and sea water
may persist for a long time span and thus bring about
substantial later dolomitization. However, this
mixing-zone model has been questioned as a viable
explanation for massive dolomites (e.g. Hardie
1987).
- Deep burial-compaction flow model (indicated in
Fig. 3.30d). Under increasing sediment loads, Mgrich pore fluids are expelled from basinal muddy sediments and may cause local dolomitization along conduits of preferential fluid flow, such as fault zones,
or partial dolomitization of platform carbonates.
However, the amount of Mg released from muddy
carbonate-rich sediments (e.g. by transformation of
high-Mg calcite to low-Mg calcite) is limited and
cannot accomplish pervasive dolomitization of thick
carbonate sequences.
- Lacustrine dolomites (Coorong model). Primary
dolomite formation is known from a number of modem lakes in regions of tropical to temperate climates
(Last 1990). One of these examples is located in
southern Australia where a large lagoon is separated
by beach ridges from several coastal plain lakes
(Coorong model, Fig. 3.30e). The lacustrine
dolomites form in shallow, saline (NaCl-rich) water
of high alkalinity and high Mg/Ca ratio (> 1 0).
The Coorong lakes are fed by Mg-rich groundwaters originating from young basaltic rocks flowing seaward through
an extensive aquifer. The groundwater mixes with sea water seepage through a seaward prograding belt of coastal
sand dunes, composed mainly of skeletal carbonate (von
der Borch and Lock 1979; Warren 1990). Strong evaporation from the shallow lakes during the dry summer season
leads to highly concentrated, Mg-rich brines and primary
precipitin of dolomite and deposition ofpellet-rich, mostly
laminated dolomitic mud. Then the lakes fall dry (mud
cracks) before being again flushed during winter time. Dolomite is associated predominantly with Mg-calcite. Dolomite makes up about 10% of the carbonate minerals in the
single process which can explain all the various phenomena observed in the rock record. Therefore several
dolomitization models have been proposed (see, e.g.,
Shukla and Baker 1988; Tucker and Wright 1990; Purser et
al. 1994; Tucker et al. 1994; Rao 1996) which are here
briefly surnmarized in a simplified way.
The various dolomitization models proposed in the
literature include:
- Sea-water contact, sea-water pumping. Normal seawater is oversaturated with respect to dolomite and
has a Mg/Ca ratio of 5.2. Consequently, sea-water is
an inexhaustible source of magnesium. It appears
reasonable to assume that at the contact of calcareous
sediment with sea-water, or pore-water of similar
chemie al composition as sea-water, part of the Ca++
of the solid calcareous phase can be exchanged for
Mg++ in sea-water. Strong wave motion and tides
pump pore-water into or out of the rock body for
very long time periods. Upward-directed terrestrial
heat flow may aid in rnaintaining rock-internal fluid
flow. The intense dolomitization of many modem
and ancient, more or less isolated reef bodies (oceanic platforms, seamounts and atolls), particularly at
their steep flanks, is usually interpreted by this
model. Similarly, the dolomitization of extensive,
thick and weB bedded shelf carbonates also seems to
require the overlying water body as Mg-source.
Downward-prograding sea water-induced dolomitization probably proceeds more or less simultaneously with the vertical upbuilding of the carbonate
sequence (bank-by-bank dolomitization).
- Brine-reflux model (or seepage-reflux model and
sabkha model with evaporative reflux). Several
dolomitization models are based on evaporation of
sea water in lagoons or coastal sabkhas in arid to
semi-arid climates. The resulting hypersaline brines
can attain a salt concentration sufficiently large for
the precipitation of gypsum (Sect. 6.4.1). Because
gypsum extracts Ca++ from the solution, the Mg/Ca
ratio of the brine far exceeds that of normal sea water
(Mg/Ca = 5.2) and therefore has the potential for initiating dolomitization of calcium carbonate in the underlying rocks. Due to its increased density, the lagoonal brine tends to sink into the underlying rocks
and then to flow back toward the sea (Fig. 3.30b). On
its way through the carbonate body, the alkaline brine
can also dissolve skeletal opaline silica which is later
reprecipitated as chert in the mixing zone between
brine and normal pore water. A slight modification of
this model is gypsum precipitation in a coastal
sabkha and brine reflux to the sea. These models are
widely applied for ancient dolomites if they are associated with evaporites.
- Meteoric-marine mixing-zone model. A number of
rock sequences with dolomites does not show any
evidence of syn-depositional or post-depositional
arid to semi-arid climate. The basic idea of the
151
mixing-zone model is the experimentally tested degree of saturation of dolomite and calcite in different
mixtures of sea water and fresh water. Normal sea
water is supersaturated for both calcite and dolomite.
When CO 2 -saturated meteoric groundwaters (95 to
50% of the solution) mix with sea water, the resulting solution becomes undersaturated for calcite, but
is still supersaturated for dolomite. Under these conditions, dolomite may form. This process is possibly
furthered by low sulfate concentrations due to the
dilution of sea water and/or bacterial sulfate reduction. In humid c1irnates, lagoonal water can become
brackish or fresh and, due to its lower density, form a
lens-shaped water body reaching into the subsurface
(Fig. 3.30c). In the mixing zone calcium carbonate is
replaced by dolomite (early dolomitization). Similarly, fresh groundwater from emerged platforms, or
from the continent can flow seaward and, dependent
on the hydraulic head, shift the mixing zone or partially flush out the primary marine pore water. The
mixing zone between meteoric water and sea water
may persist for a long time span and thus bring about
substantial later dolomitization. However, this
mixing-zone model has been questioned as a viable
explanation for massive dolomites (e.g. Hardie
1987).
- Deep burial-compaction flow model (indicated in
Fig. 3.30d). Under increasing sediment loads, Mgrich pore fluids are expelled from basinal muddy sediments and may cause local dolomitization along conduits of preferential fluid flow, such as fault zones,
or partial dolomitization of platform carbonates.
However, the amount of Mg released from muddy
carbonate-rich sediments (e.g. by transformation of
high-Mg calcite to low-Mg calcite) is limited and
cannot accomplish pervasive dolomitization of thick
carbonate sequences.
- Lacustrine dolomites (Coorong model). Primary
dolomite formation is known from a number of modem lakes in regions of tropical to temperate climates
(Last 1990). One of these examples is located in
southern Australia where a large lagoon is separated
by beach ridges from several coastal plain lakes
(Coorong model, Fig. 3.30e). The lacustrine
dolomites form in shallow, saline (NaCl-rich) water
of high alkalinity and high Mg/Ca ratio (> 1 0).
The Coorong lakes are fed by Mg-rich groundwaters originating from young basaltic rocks flowing seaward through
an extensive aquifer. The groundwater mixes with sea water seepage through a seaward prograding belt of coastal
sand dunes, composed mainly of skeletal carbonate (von
der Borch and Lock 1979; Warren 1990). Strong evaporation from the shallow lakes during the dry summer season
leads to highly concentrated, Mg-rich brines and primary
precipitin of dolomite and deposition ofpellet-rich, mostly
laminated dolomitic mud. Then the lakes fall dry (mud
cracks) before being again flushed during winter time. Dolomite is associated predominantly with Mg-calcite. Dolomite makes up about 10% of the carbonate minerals in the