Middleton G.V. (Ed.) Encyclopedia of Sediments and Sedimentary Rocks
Подождите немного. Документ загружается.
DOLOMITES AND DOLOMITIZATION
problem is that well-ordered, stoichiometrie dolomite has
never been successfully grown inorganically in laboratory
experiments at near-surface conditions of 20-30'C and
1
atm
pressure (to date, the lowest temperature/pressure at which
(proto-) dolomite has been grown under uell-controlled
laboratory conditions is 60 C/1 atm: Usdowski 1994). Hence.
geochemica! parameters that are necessary tor back-calculat-
ing the composition(s) of the dolomitizing niiid(s), such as the
equilibrium (lxygen isotope fractionation and trace element
partitioning coefficients, have to be extrapolated from high-
temperature experiments, which renders them notoriously
inacctirale and often leads to ambiguous genetic interpreta-
tions in case studies of dolomitization. Furthermore, dolomite
is fairly rare in Holocene environments and sediments, yet very
abundant in older rocks. Collectively, these aspects make up
the so-cailcd "dolomite problem", and they have led to a
number of models for dolomitization.
Chemical constraints
Dolomite formation is chemically, that is. thermodynamically
and/or kinetieally. favored by (a) low Ca^*/Mg"'-ratios: (b)
low dr'ICOJ -ratios ( = high carbonate alkalinity): (c) high
lemperatures; and (d) salinities substantially tower or higher
than that of seawater (e.g.. Machel and Mountjoy, 1986;
Arvidson and MacKenzie, 1999: Figures D32. D?3). These
constraints translate into three essential and comtnon condi-
tions and natural settings to form dolostones.
(1) A sufficient supply of Mg~' and CO? : this condition
favors marine settings and burial-diagenetic settings uiih
- 20
1.2 O.B 0.4 0.0 -0.4 -0.8
log [ca
Figure D:t2 Bivari.ilc tht'rmodyntmiu slabtlily diajiram for the system
{.ii( ilc-rlolomile-water. Square brackets denote tictivities. Lines are
talc ul.ited from experimental data: line
1
- calcite + ideal, fully
ordered dolomile: line 2 = calcite + ordered dolomite with slight Ca-
sur[jlus; lint-
.J
= calcite + fully disordered fjrotodolomite. The lour
open circles denote the results of Usdowski (1994}, whose up lo 7
year-long runs represent fhe lowest-temperature experimental
dolomite tormation performed to date. Usdownski's 11994} data for
9(1,
120 and 180 C plol close lo line 2, but his d<.ita for fiO C plols on line
'i,
whic h [irobably refiects that protodolomite rather than dolomite
formed at (i() C. Data from natural aquifers cluster close to line 2 (not
shown}, which can be considered representative of most natural
dolomite. Diagram is modified from Carpenter
(19801.
pore fluids of marine parentage because seawater is the
only common Mg-rich natural fluid in sedimentary/
diagenetic settings.
(2) A long-lasting and efficient delivery system for Mg"' and/
or CO} (also exporting Ca"' in the ease of calcite
replacement), which favors settings with an active and
iong-ktsting hydrologie drive.
(3) Carbonate depositional settings and/t)r limestones.
In addition, dolomite may form when rapidly rising
(brmation lluids suddenly reiease COi (Leach etal. 1991).
Considering thai the above chemical constraints allow
dolomite fortnation in almost the entire range of surface and
subsurface diagenetic settings, the question arises why there are
so many undolomiti/ed limestones. The es.sential conditions
for the common lack of dolomitization probably are: (d) ion
pair formation, inactivating much of the Mg''' and COi in
solution: (e) insufficient flow because of the lack of a persistent
hydraulic head, too small a hydraulic head, or insufhcient
diffusion, resulting in insufficient magnesium and/or carbonate
ion supply: (f) the limestones arc cemented and not permeable
enough: and (g) the diagenetic fluids are incapable of forming
dolomite because of kinetic inhibition, for example, because
the environment is too cold (most kinetic inhibitors of
dolomite nucieation and growth are rather potent at tempera-
tures below about $0C. and the Ca""/Mg''-ratio of many
cold diagenetic fluids is not low enough for dolomitization.
Mass balance constraints
Dolomitization is a diagenetic dissolution-replacemeni process
for which many tnass balance equations can be written.
Dolomitization in its most simplistic fonn is represented by
-- > CaMg(CO,),[s)
(reaction
+ 1.00 0.48 0,0 -0.48-1.00 -I.4B
Figure D33 Trivariale kinelic stability diagram for the system calcite-
dolomile-vvdter. The ionic rallos are molar ralios. Seawater plols just in
ihe talcile
field.
The stippleti tield boundary is bent towards higher Ca/
Mg ralios at salinities greater than i5 |)ermlL From Machel and
Mounljoy (1986}.
236
DOLOMITES AND DOLOMITIZATION
or by
CaCO.,(s) +
CaMg(CO.0 {reaction 2)
Reactions 1 and 2 are end members of a range of possible
stoichiometries. that is.
(2 - X) CaCO,(s) + Mg
CaMg{CO.0;(s) + (I - x)Ca'"(aq)
(reaction 3)
For
X
= 0, reaction 3 becomes reaction 1. and for x = I reaction
3 becomes reaction 2. Magnesium has lo be imported to and
calciutn has to be exported from the reaction site in the case of
reaction 1. whereas there is no export of calcium in the case of
reaction 2. Intermediate cases are represented by values of x
between zero and one. Dolomite cementation is most
simplistically represented by
(reaction 4)
In any given case, the amount of the diagenetic fluid needed
for dolomite formation depends on the stoichioinetry o!" the
reaction, temperature, and composition of the fluid. For
example, if dolomitization proceeds via reaction 1. about
650
m"'
of normal seawater are needed to doloinitize I
m"*
of
limestone with 40 percent initial porosity at 25' C (Land, 1985).
Dolomitization may not take place with 100 percent-efficiency,
and larger waler/roek ratios are required for complete
dolomilization by seawater in a near-surface setting. If
seawater is diluted to 10 percent of its original concentratioti.
ten times as much water is needed. On the other hand, about
30m"' of brine are needed per m' oflimestone at 100 percent-
dolomitization efficiency in (he case of a halite-saturated brine.
The role of increasing temperature in the underlying thermo-
dynamic calculations is to reduce the amount of magnesium
necessary for dolomitization because the equilibrium constant
(and hence the equilibrium Ca/Mg-ratio) is temperature-
dependent (Figure D32). For example, at 50X only about
450m' of seawater are needed for complete dolomitization of
1 m"' of limestone with 40 percent initial porosity at 100
percent-efficiency. The amounts of dilute and hypersaline
waters change accordingly.
These ealculations imply that, in all cases, large water/rock
ratios are required for cotnptete dolomitizalion. This necessi-
tates the presence of active hydrologie flow systems for
extensive, pervasive dolomitization. The exeeption are natural
environments where carbonate muds or limestones ean be
dolomiiized via diffusion of magnesium from seawater rather
than by hydrologie flow.
Porosity and permeability
Dolomitization is important for petroleum and economic
geology beeause most dolostones are more porous and
permeable than limestones. However, it is not justified to
presume that dolostones are always better hydrocarbon
reservoir roeks than limestones. Schmoker and Halley (1982)
have demonstrated with porosity/depth profiles of southern
Florida that many dolostones that have not been buried too
deeply (less than about
1
km) have porosities equal to or less
than those of adjacent limestones. A similar pattern has been
found in Devonian dolostones of western Canada that were
buried to more than 4 km (Amthor etal.. 1994). Furthermore,
the shape and extent of dolostone reservoirs are determined by
the process of dolomitization, that is, the direction and
duration of the fluid fiow regime(s). Therefore, and never-
theless, rnost dolomitized carbonates arc better reservoirs than
limestones, and dolostone reservoirs are diagenetic traps or
combination traps with a diagenetic component. A related
aspect is that dolostones often form aquifers and/or prefer-
ential migration pathways for hydrocarbons. For example, the
dolomitized margin of the Devonian Cooking Lake platform is
known as a migration pathway over a length of several
hundred km (e.g.. Amthor etal.. 1994).
Porosity may be retained, lost, or generated during dolomite
(brmation. In the case of dolomite cementation, porosity (and
permeability) loss is obvious and occurs in proportion to the
amounts of dolomite cement formed. In contrast, about
13 percent of porosity is generated, sueh that a limestone with
40 percent porosity is replaced by a dolostone with about 45
percent porosity, in the so-called "mole-per-mole" replace-
ment of calcite by dolomile aeeording to reaction 1 (whereby
two moles of ealeite are replaced by one mole of dolomite). In
such a case, the fabrics of the original limestone must be at
least partially obliterated in order to account for the volume
change during replacement.
More generally, porosity gains or losses can be represented
by the values of
x
in reaction 3. In the special cases of x = 0.11
and
X
—0.25 (for aragonite and calcite. respectively), there is
no volume loss or gain (Morrow 1982). Hence, litnestones
dolomitized in a volume-per-vohime rcplacetnent should not
contain secondary pores or dolomite cements, and the primary
textures may be partially or largely preserved. However,
partial or complete obliteration of primary textures ean oeeur
even in a volume-per-volume replacement because of a marked
change in crystal size (usually an increase, due to Ostwaldt
ripening) and/or porosity redistribution.
All the above cases of dolomitization may. and many
invariably do, involve reorganization of permeability path-
ways.
Permeability modifications are a function of the textures
and permeabilities of the precursor carbonates, as well as of
reaction kinetics during dolomitization. that is, erystal nuelea-
tion and growth rates. While porosity changes can be
estimated from the amounts of dolomite cement and/or
intercrystalline secondary porosity, permeability changes dur-
ing dolomitization are aln"iost impossible to quantify.
Dolomite textures
The most widely used classification of dolomite/dolostone
textures is the one by Sibley and Gregg (1987). based on Gregg
and Sibley (1984). This classification is popular because it is
simple and largely descriptive, although it carries genetic
implications. Crystal size distributions are classified as "un-
imodal" and "polymodar". whereas crystal boundary shapes
are classified as "planar-e"" (euhedral). "planar-s" (subhedral).
and "nonplanar"' (anhedral). Where possible, a textural
description includes recognizable allochems or biochems,
matrix, and void fillings. Particles and cements may be
unrepiaced. partially replaced, or completely replaced. Repla-
cetnent may be mimetic or nonmimetic. IJnimoda! size
DOLOMITES
AND
DOLOMITIZATION
237
distribution generally results from
a
single nucieation event
and/or
a
unimodal primary (pre-dolomite) size distribution
of
the substrate. Polymodal size distributions indicate multiple
nucieation events and/or
a
differentia! nucieation
on an
originally polymodal substrate. Planar crystal boundaries tend
to develop
at
growth below about
50 C (the
so-called "critical
roughening temperature""), whereas nonplanar boundaries
tend
to
develop
at
T>50X' and/or high degrees
of
super-
saturation.
All
exception
are the
planar
yet
curved crystal faces
of saddle dolomite, which
is
known
to
form only
at
elevated
temperatures
of
more than about
80 C
(Radtke
and
Mathis,
1980:
Machel, 1987). Saddle dolomite
is
different also
in
other
respects,
for
example,
it
almost invariably
is
sparry
and
commonly milky-white,
it has
distinctively warped crystal
lattices, undulose extinction, pearly luster,
and
abundant fiuid
inclusions.
Textures
and
reservoir characteristics
of
natural dolostones
arc highly variable. Where dolomitization
is
matrix-selective,
significant porosity tends
to be
remnant primary porosity with
relatively poor intereonneetion. Undolomitized calcite fossils
commonly
are
dissolved during
or
after dolomitization.
leaving
a
rock resembling Swiss eheese, that
is, a
relatively
tight matrix with large holes. Various types
of
eement. often
anhydrite, partially
ov
completely occlude such secondary
porosity. Where dolomitization
is not
matrix-selective,
or
where
the
matrix
is
relatively coarse-grained, sucrosic dolos-
tones with intercrystalline porosity
and
excellent permeability
tend
to
form.
Dolomite geochemistry
In addition
to
textures
and
their distribution relative
to
stratigraphy
and
structure,
a
wide range
of
geoehemical
methods
may be
used
to
characterize dolomites
and
dolos-
tones.
and to
decipher their origin.
The
most widely used
methods
are
stable isotopes (O. C), Sr-isotopes, trace elements,
fiuid inclusions,
and
paleomagnetics (e.g.. various case studies
in Purser vial.. 1994). Care must
be
taken
to
assess
the
degree
of recrystallization, which
may
result
in
dolomite compositions
that
are
characteristic
of the
environment(s)
of
recrystalliza-
tion rather than
the
original environment(s)
of
dolomitization
(e.g.. Machel, 1997).
Natural environments
and
models
of
dolomitization
Dolomite occurs
in
many diagenetie environments that range
from
the
surface
to
deep subsurface settings
of
several
km
burial depth. Most dolostones originate
by the
replacement
of
limestones that form(ed)
in
shallow marine environments with
normal seawater salinity, which
is
shown
by
relics
of
primary
sedimentary features, such
as
ripple marks, reef fossils,
burrows,
etc.,
that have survived
the
replacement process.
In
the Recent, however, dolomite
is
almost absent from such
carbonate depositional environments, despite
the
faet that
seawater
is
many times supersaturated with respect
to
dolomite. Thtis. various models
nj
elolt>miti~ation
have been
devised
to
explain
the
origin
of
these types
of
replacement
doiostones.
and
almost every time
a new
model
has
been
proposed,
it
became
a
bandwagon until
the
next popular
model arose. There also
are
dolostones that
are
devoid
of
primary sedimentary features,
and
where
an
origin without
a
limestone precursor appears possible. Dolomite
is
also
common, albeit generally
not
abundant,
as a
cement
in
limestones
and
clastic rocks.
The potential
of
natural environments
to
form dolomite
and
dolostone
can be
assessed
on the
basis
of the
above chemical
considerations,
and
using hundreds
of
case studies that have
been published over
the
last five decades
(see
compilations
in
Zenger
cl al.. 1980:
Shukia
and
Baker,
1988:
Purser
et
ciL.
1994).
Genetically,
all
natural dolomites
can be
grouped into
two families (Budd. 1997). Pemronlemporaneoiis dolomites
form while
the
host sediments
are in
their original depositional
setting. Most known penecontemporaneous dolomites
are of
Holoeene
age.
However, there probably
are
many older
examples
in the
geologic record,
yet
they
are
rnuch more
difficult
to
prove. Postdepositionaldolomile.s form after deposi-
tion
has
ceased
and the
host carbonates have been removed
from
the
zone
of
active sedimentation
by
progradation
of the
depositional interface, burial, uplift, eustatic sea-level change,
or
any
combination
of
these factors. Almost
all
dolostones
are
postdepositional. Additionally, small amounts
of
dolomite
may
be
formed syndepositionally
or
postdepositionally
by the
magnesium derived from within
the
dolomitized sediments,
that
is.
originally contained
in or
adsorbed
by
Mg-calcites.
organic matter, chlorophyll, biogenic silica, certain clay
minerals,
or
older detrital dolomites,
if
present (e.g., Lyons
etal.. 1984; Baker
and
Burns, 1985).
Penecontemporaneous dolomites and
the
microbial/
organogenic model
There
are two
preferred settings
of
penecontemporaneous
dolomite formation, that
is.
shallow-marine
to
supratidal,
and
hemipelagic
to
pelagic.
In
either case,
and
with
few
exceptions,
the amounts
of
dolomite formed
are
small,
and
these dolomites
are texturally
as
well
as
geochemieally distinct.
Penecontemporaneous dolomites
in
shallow-marine
to
supratidal settings usually form
in
quantities
of <5 vol-
percent, mostly
as
Ca-rich
and
poorly ordered, fine-crystalline
eements and/or directly from aqueous solution. One especially
important type
of
peneeontemporaneous (proto)dolomite
[brms lenses
and
layers
of up to 100
percent
in
sabkhas.
as
discussed further below.
The
other occurrences
are as
lithiHed supratidal crusts (e.g., Andros Island; Sugarloaf
Key; Ambergris Cay);
as
thin layers
in
salinas (e.g., Bonaire;
West Caicos Island)
and
evaporative lagoons/lakes (e.g.,
Coorong):
as
fine-crystalline cements
and
replacements
in
peritidal sediments (e.g.. Florida
Bay:
Andros Island):
and as
fine-crystalline supratidal weathering product
of
basic rocks
(Capo etal.. 2000: this
is the
only known case
to
date where
penecontemporaneous dolomite
is
ordered
and
stoichio-
metrie).
The
dolomite-forming fluid
is
normal seawatei"
and/
or evaporated seawater. rarely with admixtures
of
evaporated
groundwater.
Penecontemporaneous dolomite formation
in
hcmipelagic
to pelagic settings also form very small quantities
of
microcrystalline protodolomite.
if any.
that
is,
generally less
than
I
weijzht-peccent (Lumsden. 1988). However, under
favorable circumstances
the
amount
of
dolomite
can
reach
up
to 100
percent locally.
For
example, Miocene hemipelagic
carbonate sediments from
the
margin
of the
Great Bahama
Bank
are
partially
to
completely dolomitized over
a
depth
of
range ofabout 50-500
m
subsea. Dolomite occurs
as a
primary
void-filling cement
and by
replacing micritic sediments,
red
algae,
and
echinoderm grains (Swart
and
Melim, 2000).
238 DOinMITFS AND DOLOMITIZATION
Both settings of peiieconlemporaneous dolomite formation
appear to be linked to the "microbial model'" or "organogenic
model" of dolomiti/ation (Vasconcetos and McKenzie. 1997;
Burns elal., 2()()(); Mazzullo, 20(10). Accoi'ding to this model,
dolomite may be formed syndepositionally or early-postdepo-
sitionally at depths ofa few cm lo a few hundred meiers under
the intliience of ox promoted by bacterial sulfate reduction
and/or mcthanoy;cncsis, as shown by iVC-values, The specific
role of microbial activity in reducing the kinetic notorious
barriers to dolomitization is presently uncertain, although
reduction of Mg- and Ca-hydration barriers, an increase in
alkalinity and changes in pH, seem to be involved. Most
microbial/organogenic dolomites are cements, but some are
replacive. They typically arc finc-cryslalline (less than 10).im).
calcic and poorly ordered protodolomites. The chief modes of
Mg-supply are diffusion from the overlying seawater and/or
release from Mg-calcites and clay minerals. Microbial/organo-
genic dolomites may act as nuclei for later, more pervasive
dolomitization during burial.
Hyposaline environments and the mixing zone model
Hyposaline environments arc those with salinities below that
of normal seawater (35 mg/i). This includes coastal and inland
freshwater/seawater mixing zones, marshes, rivers, lakes, and
caves.
Postdepositional dolomite has been found to form in all
o^
these environments, but only in small amounts and
commonly as cements.
One hyposaliue environment, the coastal freshwater/sea-
water mixing zone (often simply called mi.xing zone) has given
rise to one of the oldest and most popular models, that is. the
"mixing zone model'". However, the mixing zone model has
been consistently overrated since it was first proposed
(Badiozamani. 1973: Land, 1973). For example, Humphrey
and Quinn (1989) suggested that coastal mixing zones may
form thick sections of dolomite in a pi at form-margin setting,
and that such dolostones may be common in the geologic
record. Other authors have proposed mixing zone dolomitiza-
tion tbr entire carbonate platforms many hundreds of knr in
size (e.g., Xun and Fairchild. 1987).
Dolomite may form in freshwater/seawater mixing zones.
This is due to the fact that many freshwater/seawater mixing
zones are thermodynamically undersaturated for calcium
carbonate yet supersaturated with respect to dolomite in at
least a part of the mixing speetruni (commonly in mixtures of
about 10 5U percent seawater). However, not a single location
in the world has been shown to be extensively dolomitized in a
freshwater/seawater mixing zone, in recent or in ancient
carbonates. Rather, dolomite formation appears to be
kineticaliy inhibited in the vast majority of coastal mixing
zones (Smart el al., 1988: Machel and Mountjoy, 1990),
Furthermore, calcite and aragonite dissolve at rates many
times faster than dolomite is formed, thus, the most common
result of mixing zone diagenesis is rocks riddled with
secondary porosity up to the dimensions of caves, with or
without small amounts of dolomite (e.g., Florida. Yucatan:
Baek et al., 1986: confirmed by modeling: Santord and
Konikow. 1989). In addition, most coastal mixing zones are
only a few hundreds of meters wide and pass relatively quickly
through the rocks in response to custatic sea level fluctuations
and/or subsidence. Hence, even where mixing zones arc
capable of forming dolomite, the dolomitized rock volume is
relatively small and restricted to the plattbnn margin(s)
(Figure D34(A)).
Most mixing zone dolomite is petrologieally and geochemi-
cally distinct. Most crystals are relatively clear, planar-e or
planar-s. stoichiometric, well-ordered rhombs. However, some
mixing zone dolomite is protodolomite. Crystal sizes eom-
monly range from I micron to 100 microns but reach several
millimeters in some cases. Most mixing /one dolomite occurs
as cements in microscopic mterstices and macroscopic voids,
molds, vugs, and caverns. Alternating generations or growth
zones of ealcite/dolomite are common in coastal mixing zones
with rapid and cyclical changes of salinity.
The main role of eoastal mixing zones in dolomitization
might be thai ofa hydrologie pump for seawater dolomitiza-
tion. rather than that of a geoehemical environment favorable
for dolomitization (Machel and Mountjoy. 1990). At the
seaward margin of a eoastal mixing zone seawater is pumped
HydfOlogic nx)del
Predicted dolomite patterns
A Coastal mixing zone
A ^^ A
B Reflux
C Compaction-driven flov/
D Slope convection
E Slope convection
F Regional-scale flow
Figure D34 Sck'Lled mudels ol" dolomitization, illustr<ited as
groundwater flow systems (left) and predicted dolomitization patterns
(right), Ex,iniplfs <ire uf incomplete dolomitiztition of tarbonate
platforms or reefs. Arrows denote flow directions: dashed lines show
isotherms. Predicted dolomitization patterns are shaded. Models A-E
are km scale; model F is basin-scale. Reproduced from Amthor ef jl.
DOLOMITES AND DOLOMITIZATION
239
through the scdimcnss by the hydrologie action of tVcshwiitcr
scjuatcr mixing, whereby the water movemeni is influenced
also by tidal pumping. Hence, substantial amounts of seawater
may be pumped through the affected sediments by the
hydrologie forces of freshwater seawaler mixing and the tides.
Hypersaline environments and the reflux model
Intertidal to supratidal environments with salinities greater
than that of normal seawater occur throughout the world at
latitudes of less ihan about 30^ Extensive, pervasive dolomi-
ti/ation of shallow water carbonates, up to the dimensions of
huge and thick carbonate platforms, from density-driven ilou
(retlux) of the associated evaporitic brines ha^ been suggested
b\
many authors on the basis oi' ease studies and modeling
(e.g.. Shields and Brady. 1995). This "reflux model"
(Figure D34(B)) has been another very popular model of
dolomitization throughout the last 30 years.
The sabkha of the Trucial Coast of Abu Dhabi, the
probably best researched recent hypersaline intertidal lo
supratidal flat (Patterson and Kinsman. 1982; Baltzer elal..
1994).
is representative of prolific reflux dolomite formation,
I'ine-cryslallinc protodolomile is formed penccontempora-
neously by evaporating tloodwaters that are propelled
periodically t)nto the lower supratidal zone and along remnant
tidal channels by strong onshore whids. Dolomite formation
takes place in a narrow (1-1.5 km) fringe next to the strandline
and within the channels within the upper I m to 2ni of the
sediments, and preferentially where the environment is
chemically reducing (sulfate reduction, enhanced carbonate
alkalinity). The distribution of protodolomite is uneven across
the sabkha. and the best dolomiti/ed parts eontain 5 to about
65 weight pereent protodolomite. Dolomite forms as a cement
and aragonite is replaced, but lithification does not oecur. or
only partially. A variety ol" evaporite minerals, most notably
multigenetic calcium suifates, form within the dolomitized part
oi'
the sabkha.
In most respeets. the Abu Dhabi sabkha appears to be a
good recent analog for dolomitization in many ancient
intertidal to supratidal ilals. such as landward of the famous
Permian carbonates of Texas and New Mexico. Rather than
forming reservoir rocks, such doiostones- including the
associated evaporiies—generally form tight seals for under-
lying hydrocarbon reservoirs (e.g.. Major elal.. 1988).
In general, sabkhas and 'similar' intertidal to supratidal
environments in more humid climates typieally fonn small
quantities of fine erystalline protodolomite in thin beds, crusts,
or nodules, either within the upper I m to 2 m of sediment, or
at the sediment surface. Repeated transgressions and regres-
sions may stack such sequences upon one another. Further-
more, it appears that the efficiency of reilux is different during
transgression and regression, Mutti and Simo (1994) i'ound in
Permian carbonates that dolomitization occurred during
transgressive cycles only in intcrtidal and supratidal facies.
whereas dolomitization affected supra-, inter-, and subtidal
i'acies during regression. They also speculated that tidal
pumping and evaporative pumping may aid reflux in supplying
Mg. Ill contrast. Montanez (1997) ibund in Ordovician
carbonates that transgressive cycles were not dolomitized by
reflLix. whereas the facies within regressive cycles were almost
completely replaced by tight finc-crystallinc reflux dolomite.
He this as it may. deep penetration by reiluxing evaporative
brines and/or platibrm-wide dolomitization has not been
demonstrated anywhere in the world.
On the other hand, some the latest sophisticated modeling
suggests that refluxing brines originating on 'permanently'
flooded carbonate platibrms may penetrate the platforms
rather deeply, that is. I-2 km (Jones
['/(//..
2()02a.b). This
situation difi'ers from sabkhas and similar intertidal to
supratidal Hats, as the latter are flooded only intermittently.
Even in flooded carbitnate platforms, however, dolomitization
is predominant and/or restricted to the upper few meters or
tens of meters. Below these depths, the fluxes are too low and
the brines, although still refluxing, have exhausted their
dolomitization potential. Modeling the Devonian Grosmont
platform in western Canada, based on extensive stratigraphie.
lithologic. and geoehemical data. Jones ei at. (2{)()2b) found
thai after 100
k
years, during which brines concentrated up to
150 per mil (slightly beyond gypsum saturation) overlay the
platibrm. reflux develops and shuts oiTgeothennal circulation.
The onset of reflux results in a reversal in groundwater flow
direction as brines are transported irom the platibrm interior
basinwards. Refluxing brines enter the platform in a zone that
extends approximately 150 km from the coast. As the platform
evolves the brine plume, which is preferentially orientated in a
horizontal direction, sinks into the plati'orm subsurfaee and
migrates basinwards. Computed distributions of fluid flux in
eonjunction wilh Mg mass-balance calculations permit the
tracking of dolomite fronts as the Grosmont plati'orm evolves.
Within the range of parameter uncertainty, reflux could have
delivered sufficient Mg in 1.6 m years to account for the
pervasive dolomitization of the Grosmont Formation. How-
ever, even in these modeling results the amount of dolomite
ibrnied is not 100 percent, and doiomitization is restrieted to
layers and beds of dolomite that arc interrupted by limestones,
which results from repeated transgressions and regressions.
A previously unrecognized variant of reflux, proposed on
the basis of numerical modeling by Jones el al. {2002a), is
"'latent reflux". This type of reflux describes the phenomenon
whereby brines of reflux origin continue to sink into a piattbrm
after the cessation of brine generation on the platform top.
Hence, latent reflux could deliver magnesium to the plati'drm
carbonates after the cessation of evaporation, thereby ibrming
additional dolomite. However, latent reflux, as active reflux
during e\aporation. would exit along or near the platform
margin, and the brines would not lead to any dolomitization at
greater depths. The efficacy of this process with respect to
dolomitization remains to be proven,
Seawater dolomitization model
Postdepositional formation of massive dolostones has been
attributed to the "seawater dolomiti/ation model" (Purser
etal.. 1994). with specific rcicrence to the Cenozoie dolostones
ofthe Bahama platform, which are oiten used as an analog for
older dolomitized carbonate platforms elsewhere. Various
hydrologie systems been invoked to drive large amounts of
seawater through the Bahama platform, that is. thermal
convection (Sanford ei ul.. 1998); a combination of thermal
seawater convection and reflux of slightly evaporated seawater
derived from above (Whitaker ciul.. 1994); or seawater driven
by an overlying i'resh water/sea water mi.xing zone during
partial platform exposure {Vahrenkamp and Swart. 1994),
possibly layer-by-tayer in several episodes (Vahrenkamp etal.,
1991).
All these studies agree thai the dolomitizing lluid was
240 DOLOMITES AND DOLOMITIZATION
essentially seawater that was. at best, only slightly altered by
water-rock interaction.
it is noteworthy that the Bahama doiostones represent a
hybrid regarding classification. This is because the dolomitiz-
ing solution probably was nearly normal seawater. yet thermal
convection, as a hydrologic system and drive for dolomitiza-
tion. is better classified under burial (subsurface) models.
Similarly. Devonian dolostones in western Canada, which
probably formed at depths of 300
1,500m
irom slightly
altered seawater. have been classified as burial dolostones
(Amthor etal., 1994), yet they could equally well be classified
as seawater dolostones. This classification dilemma arose from
the historical evolution of our understanding of these
dolostones.
Burial (subsurface) environments and related models
Burial (subsurface) environments are those removed irom
active sedimentation by burial, and where the pore fluid
chemistry is no longer entirely governed by surface processes,
that is, where water rock interaction has modified the original
pore waters to a significantly degree (whereby "significant" is ill
defined). All burial (subsurface) models for dolomitization
essentially are hydrologic models that difi'er mainly in the
hydrologic drives and direetion(s) of fluid flow (e,g,. Morrow,
1999).
Four main types of fluid ilow take place in subsurface
diagenetic settings: (I)compaction flow: (2) topography-driven
flow: (3) thermal convection: and (4) tectonically-driven flow.
Mixtures of these flow regimes and fluids are possible under
certain circumstances.
The oldest burial model of dolomitization is the "eompae-
tion model" (Illing. 1959: Jodry, 1969). In this model (Figure
D34(C)). seawater and/or its subsurface derivative is pumped
through the roeks at several tens to several hundreds of meters.
The model was never especially popular because it became
clear that burial compaction can only generate fairly limited
amounts of dolostone due to the limited amounts of
compaction water.
Convection through carbonate platibrms may be driven by
heat from below or by density differences from above. One
type of thermal convection originates from the horizontal
density gradient caused by the regional temperature difference
between the interior of uplifted platibrms and the surrounding
seawater. In this case, convection may take plaee in half-eells
that are open to the adjacent and overlying ocean, with
marked differenees in the fluxes and heat transport as a
function of the ratios of horizontal to vertical permeability
(Figures D.34(D). (F): Santbrd etal.. 1998). The amounts of
dolomite that ean be tbrmed in such conveetion half-cells are
essentially unlimited, that is. dolomite can be formed as long as
convection takes place. In contrast, the amounts of dolomite
are much smaller where convection takes plaee in full-cells that
are not open to the ocean. This may happen where convection
is driven by a relatively weak regional heat gradient or by a
loeal heat source, such as a magmatic intrusion. Thermal
convection has been suggested for extensive, pervasive
dolomitization in several carbonate platforms and/or regional
carbonate aquifers (e.g.. Morrow, 1999). ln partieular. thermal
convection has been advoeated as the chief mode of
dolomitization fbr the Cenozoie carbonates of the Bahamas
platform (here discussed under "Seawater dolomitization
model").
AMBtENT
TEMPBtATURE
HYDROTHERMAL GEOTHERHAL
HYDROFRIGtD
200 C
Figure D35 Hydrothermal, geothermal {formed in thermal equilibrium
with the surrounding rocks), and hydrofrigid mineral formation, rrum
and Lonne |2OO2).
Convection cells invariably have rising limbs that penetrate
the overlying and cooler strata, linking thermal convection to
hydrothermal dolomitization. There are well-documented
examples of hydrothermal dolomite, in local as well as in
regional flow systems (e.g.. Spencer-Cervato and Mullis, 1992:
Qing and Mountjoy 1992, 1994; Duggan ei al., 2001).
However, many dolomites that formed at elevated tempera-
tures,
especially but not exclusively saddle dolomites that
fonned at T>80"C, have mistakenly been called hydrotherma!
(e.g.. Davies. 1997), The term "hydrothermal" is and should be
Lised aeeording to White's (1957) time-honored definition as
"aqueous solutions that are warm or hot relative to the
surrounding en\ironment". with no genetic implications
regarding the fiuid source, including Stearns et (tl.'s (1935)
rationale for a "significant" temperature difi'erence of at least
5 IOC (Figure D35), Henee. any mineral, including dolomite,
should be called "hydrothermal" only if it can be demonstrated
to have ibrmed at a higher (by >5-10 C) than ambient
temperature, regardless of fluid source or drive (Figure D35.
left).
This definition does not carry a lower or upper
temperature limit, A dolomite formed at only 40 C could be
hydrothermal. if the surrounding rocks were significantly
colder than that at the time of dolomite formation, ff
a
mineral
was formed at or near the same temperature as the
surrounding rocks (within 5- lO^'C), it should be called
"geoihermal" (Figure D35. center), whatever the geothermal
gradient, ln keeping with common usage, the qualiiier
"geothermal" may simply be omitted, unless speeial emphasis
needs to be placed on the geothermal nature of a particular
mineralization event. Minerals ibrmed at temperatures sig-
nificantly lower than ambient (by >5-]0 C) may be called
"lmlfojrif>icl".
even if they formed at a rather high temperature
(Figure D35. right). For example, seawater or freshwater may
penetrate a rock sequence through highly permeable pathways,
such that the water is heated to 150X at a depth where the
surrounding rock has a temperature of 250 C, If a mineral
were formed at that time from the incompletely heated water,
the mineral would be hydrofrigid.
Topography-driven flow takes plaee in uplifted sedimentary
basins that are exposed to meteoric recharge. With time
enormous quantities of meteorie water may be pumped
through a basin, often concentrated by water-rock interaction
(especially salt dissolution), and prei^erentially funnelled
DOLOMITES AND DOLOMITIZATION
241
through aquifers. Significant dnlomitization can only take
place, however, where the meteoric water dissolves enough Mg
en route before encountering limestones. Gregg (1985)
advocated topography-driven flow for regionally extensive
dolomitization in Cambrian carbonates of Missouri.
Another type of tlow that has been suggested to result in
pervasive dolomitization is teetonieally-driven. that is, meta-
morphic fluids expelled from crustal sections affected by
tectonic loading that also drive basinai fluids toward the basin
margin (Figure D34(F)). Sueh fluid flow could be injected into
compactional and/or topography-driven flow, with attendant
fluid mixing. Dorobek (1989). Drivet and Mountjoy (1997),
and others have invoked tectonically-driven flow for dolomi-
tization of the Siluro-Devonian Helderberg Group. USA, and
for Devonian reefs in Canada, respectively. However, it is
hypothetical at the present time whether lecionically induced
flow, or the related lluid mixing, can form massive dolostones.
because the fluxes of teetonically induced flow appear to be
low (e.g., Machel etat., 2000).
Textures and reservoir characteristics of dolostones fbrmed
in burial (subsurface) environments are variable. Where
dolomitization is matrix-selective, significant porosity tends
to be remnant primary porosity with relatively poor inter-
connection. LJndolomitized caleite tbssils may be dissolved
during or after doiomiti/ation. leaving molds and \ugs. Where
dolomitization is not matrix-selective, or where the matrix is
relatively coarse-grained, sucrosic dolostones with intererystal-
line porosity and excellent permeability tend to be formed.
Unfortunately, these textures are not indicative of any of the
genetic possibilities discussed above, with the exception of
saddle dolomite, which appears to form only at T>80 C. This
dolomite is texturally and geochemically distinct. Saddle
dolomite occurs as cements in voids, as replacive masses
(often as part of MVT-suliide mineralization), or as so-called
zebra dolomite in alternating layers with much tiner-crystalline
matrix dolomites. Zebra dolomite is much debated and may
originate from repeated fracturing or some type of geoehem-
ical self-organization (Krug cial.. 1996).
Secular distribution of dolostones
The relati\e abundance of dolostones thai originated from the
replacement of marine limestones appears to have varied
eyclically through time, with two discrete maxima of "sig-
nificant early"" dolomite ibrmation (massive early diagenelic
replacement of marine limestones) during the Phanerozoic.
that is. the Early Ordovieian/Middic Silurian and the Early
Cretaceous (Given and Wilkinson. 1987). Various explana-
tions have been proposed tbr this phenomenon, that is. that
periods of enhanced early dolomite formation were related to
or controlled by plate tectonics and related changes in the
compositions of the atmosphere and seawater. such as an
increased atmospheric C02-level. high eustatic sea level, low
saturation state of seawater with respect to calcite, ehanges in
the marine Mg/Ca ratio, or low atmospheric O^-levels that
coincided uith enhaneed rates of bacterial sulfate reduction
(e.g., Stanley and Hardie. 1999: Burns etal.. 2000). At present.
the reason(s) for the alleged or observed secular variations in
dolostone abundance remain(s) uncertain.
Concluding remarks
There are many environments and possibilities to form
dolomites and dolostones. However, the (paleo-) hydrologic
characteristics are rarely known well enough for an unambig-
uous interpretation of the dolomitizing fluid How regime.
Dolostones often eonsist of different generations of dolomite
formed by different flow systems and mechanisms. Metastable
(proto-) dolomites may be formed, especially in near-surface
and shallow burial settings, which later and deeper recrystallize
to more stable dolomites, either in one or in several
recrystallization steps.
Mans G. Machel
Bibliography
Amthor, J.E.. Mountjoy. E.W.. and Machd. H.G., 1994. Regional-
scale porosily and permeability variations in Upper Devonian
Led lie
buildups: implications for reservoir developmeni and
prediction in carboniites. Anieriiaii
A.ssiHiiitioii
fetiolvum (ji'ntn-
gtstsButtetiii. 78: 1541
\^^'-).
Arvidson. R.S.. and MaeKenzie. E.T-, 1999. The dolomite problem:
control of preeipilalion kinelies by temperature and saturation
state.
Ameriai'i
J«urii(iloJ
Scie/Jie. 299: 257 28X.
Badiozamani. K..
197.1.
The Dorag dolomiiizalion model appliealion
to the Middle Ordovieian of Wisconsin. Jtiiinuit of Seilimenleirv
Petrology. 43: 965-984.
Back. W.. Hanshitw, D.B.. Herman. J.S.. and van Drid. J.N.. 1986.
Differential dissolution ofa Pleistocene reef in the groundwater-
mixing/one ol'eoasta! Yucatan, Me.Kico. Geology. 14: 137 140.
Baker. P.A.. and Burns, S.J.. 1985. Oeciirreiice and formation of
dolomite in organic-rieh continental margin sediments. .4iin'rietiii
Aysot iitlian Petniletim (k'(>l(>i!iM.\ Bitllctiii. 69: 1917- 1930.
Baltzer. E., Kenig, F.. Boietiard. R.. Plazial, J.-C. and Purser, B.H.,
1*^94.
Organic maiter distribution, water circulation and dolomi-
tization beneath llie Abu Dhabi sabhka (LInited Arab Emirates). In
Purser. B.. Tucker. M.. and Zenger. D, (eds.), Dutomites—A
\uliiinf in Honor of Dutoinieii. International Assoeiation of
SedimL'riiologists. Special Publication. 21. pp. 409 428.
Budd. D.A.. 1997. Cenozoie dolomites of earbonate islands: their
iiUribulcs anti origin. Eurth-Svii'iue Reviews. 42: 1-47,
Burns. S,J., McKenzie, J.A.. and Vaseoncelos, C. 2()()(). Dolomite
Ibrmiition and bioclietnical cycles in the Phanerozoic. Sediineiirol-
<'^'v. 47:
49-61.
Capo.
R.C, Whipkcy. C.E.. Biaehere. J.R.. and Chadwiek. O.A..
2(H)().
Pedogenic origin of dolomite in a basaltic weathering profile,
Koliala Peninsula, Hawaii. Ceoloi-y. 28: 271 274.
Carpenter. A.B.. 1980. The ehcmislry of dolomite Ibrmalion I: the
stability of dolomite. In Zenger. D.H.. Dunham. J.B.. and
Ethington. J.B. (eds.). Cmuept.K anil Mailels o/ Doloniitizaiion.
Society of Eeonomic Paleontologists imd Mineralogists Spt:cial
Publieation. 28. pp. Ill 121.
Davies. G.R., 1997. Hydrothermal dolomite (HTD) reservoir facies:
global perspectives on tectonic-slruetiiral and temporal linkage
beiween MVT and Sedex Pb-Zn ore bodies, and subsurfaee HTD
reservoir facies, Camiilian .Soeiely oj Petvolciiin Geoloiiists Short
Cour.se
Notes.
167 p.
Drivet. E.. and Mountjoy. E.W.. 1997. Dolomitization of tbe Leduc
Eormation (Upper Devonianl. southern Rimbey-Meadowbrook
reef trend. Alberta. Journal of Sediiuenuirv Research. 67: 411 423.
Duggan. J.P., Mountjoy. E.W., and Stasiuk. L.D.. 2001, Eault-
eontrolled dolomitizalion at Swan Hills Simonette oil field
(Devonian), deep basin west-eentral Alberta. Canada. Seiliiiicntol-
o^y. 48: 301 323,
Dorobek. S..
1*^89.
Migration of orogcnic fluids tlirougli the Siluro-
Devonian Holderbcrg Group during late Paleozoic deformation:
constraints on lluid sources and implications for thermal histories
of sedimeniary basins,
Ti\u>n<iphy.\ic.s.
159: 25-45.
Given. R,K,. and Wilkinson. B,H., 1987, Dolomile abundance and
stratigraphic age: eonstrains on rates and meehanisms of Plianer-
ozoic dolostone formation. Journal of Setlimenltiry
Petrolusiv.
57:
1068 1078,
Gregg. J,M,. 1985. Regional epigenetic dolomitization in the
Bonneterre dolomite (Cambrian), southeastern Missouri. Geoliigv,
13:
503 506.
242
DOLOMITCS AND DOIOMITIZATION
Gregg. J.M.. and Sibley, D.F., 1484. Epigenetie dolomitization and the
origin ot" xenotopic dolomite texture, .linivna! of Seilinicnhirv
Pvlroloi^y.
54: yOX-tJ.H,
Humphrey. J,D,, and Quinn. T,M,. 1989, Coastal mixing zone
dolomite, forward modeling, and massive dolomitization of plal-
fortn-marain carbonates. Journal of Scilinienlarv Pflrolof^y. 59:
438 454."
Illing, L.V.. 1959. Deposition and diagenesis of some upper Paleozoic
carbonate sediments in western Canada, Fifth
Worhl Peirolvtim
Con-
gress New
York.
Proceedings
Seclion I. pp, 23-52.
Jodry. R,L,. 1969, Growlh and dolomilization of Silurian Reefs.
St. Clair County. Michigan, Ameriean As.soeiation Petroleum
Geolo^^isl.s
Bulletin. S3: 957 981.
Jones.
G.D.. Whitaker. F.F,. Smart. P.L.. and Sanlord. W,E,. 2()02a.
Fate of relliix brines in carbonate platforms.
Gt'olo,i;y
30: 371 374.
Jones.
G,D,. Smart. P,L.. Whitaker. F,F,. Roslroti. B,J.. and
Machel. H,G,, 2(K)2b. Numerical modeling of lellux dolotnttization
in the Grosmont Platform eomplex (llpper Devonian). Western
Canada Sedimentary Basin. .Ann-rieini As.soi-ialion of Petroleum
Geologisis Bullelin. submitted,
Krug. H.-J.. Biandstadter, H.. and Jaeob. K.H.. 1996, Morphological
instabilities in pattern formation by precipitation and crystal-
lization proeesses. Geologisehe Runilsehau. 85: 19 28,
Land, L,S,. 1973. Contemporancons dolomiti/ation of middle
Pleistocene reefs by meteoric water. North Jamaiea, Bullelin Min-
ine Seienee. 23: 64-92.
Land, L.S.. 1985, The origin
o!"
massive dolomite, .lournalofGeoiogieul
Eihuation. 33: 112 125.
Lasemi, Z,, Boardman. M,R.. and Sandberg. P.A.. 1989. Cement
origin of stipraiidal dolomite. Andros Island. Bahamas, Journalof
Sedimenlarv
Pctrolo^-v.
59: 249 257,
Leaeh. D.L,. Plumlee, G,S,. Hofstra. A,H,. Landis. G,P,. Rowan. E,L,.
and Viets. J,G,. 1991, Origin of late dolomite cement by COi-
saturatcd deep basin brines: evidence trom the O/ark region,
central United States, Geology. 19:
348-351,
Lumsden. D,N,. 1988, Characteristics of deep-marine dolotiiites,
Journalof Seitinwntary Peirohgy. 58: 1023-1031.
Lyons. W,B,. Hines. M.E.. and Gatidette, H.E,. 1984, Major and
minor element pore water geochemistry of modern marine sabkhas:
the intluence of cyanobacterial mats. In Cohen. Y.. Castenhob.
R,W., and Halvorson. H,0. (eds,). Microhial Mais. New York: A,
R, Liss. pp, 411 423,
Machel. H.G.. I9!^7, Saddle dolomite as a by-product of chemical
compaction and thermochemical sttlfate reduction. Geoloi-v. 15:
936 940.
Machel.
H.CJ..
1997, Recrystaliization versus neomorphism. and the
concept of "signilicant recrystalli/ation" in dolotnile research. Setli-
nieniary
Geology.
113: 161-168.
Maehel. H,G,, and Lonne. J,. 2002. Hydrothermal dolomite—a
product of poor definition and itiiagination, SedimenKirv Geology.
152:
163 171,
Maehel. H,G,. and Mountjoy, E,W,, 1986, Chemistry and environ-
ments of dolomitization—a reappraisal. Earth Seienee Reviews. 23:
175 222.
Machel. H,G,. and Mountjoy. E,W,, 1990, Coastal mixing zone
dolomite, forward modeling, and mtissive dolomiti/ation of p!at-
form-martiin carbonates Discussion, Journal of Sedimeniary Pet-
rology. 601 1008-1012,
Machel. H.G,. Mountjoy. E,W,. and Amthor. J.E,. 1996, Mass balance
and lluid How constraints on regional-scale dolomitization. Late
Devonian. Western Canada Sedimeniary Basin. Bullelin of Cana-
dian Petroleum Geology. 44: 566 571.
Machel. H,G,. Cavell. P..\.. Buschkuehle, B,E,. and Michael, K.. 2000,
Tectonieally induced tluid flow in Devonian carbonate aquifers of
the Western Canada Sedimentary Basin. Journal of Geoeheniieal
Exploration. 69-70: 213-217,
Major. R.P,. Bebout. D,G.. and Lucia, F,J,. 1988. Depositional facies
and porosity distribution. Permian (GuadalupianI San Andres and
Grayburg Formations. PJWDM Filed Cotnplex, Central Basin
Platform. West Texas, In Lomando. A,J,. and Harris. P.M, (eds,).
Giani Oil anil
Gas Fields—A Core
Workshop.
SEPM Core Workshop
12.
pp. 615-648.
Mazzullo, S,J.. 2000, Organogenic dolomiti/ation in perilidal to
deepsea sediments. Journalof Sedimeniary Research. 70:
10-23,
Montane/. 1,P,. 1997, Seeondary porosity and late diagenetic cements
in the Upper Knox Group, central Tennessee region: a tetnporal
and spatial history of lluid How conduit development uithin tbe
Knox regional aquifer. In Montane/. I.P,. Gregg. J,M.. and
Shelton. K,L,. (eds,). Basin-Wide Diagenelic Patterns: Integrated
Pelrologic. Geoehemical. and Hydrologie Consideralions. SEPM
Speeial Publication. 57. pp. 101-117,
Morrow. D.W.. 1982, Diagenesis II. Dolomite—part II: Dolomiliza-
tion models and ancient doloslones. GeoseienceCanada. 9: 95-107.
Morrow. D,W.. 1999, Regional subsurface dolomiti/ation: models and
constraints.
Gco.seience
Canada, 25: 57 -70.
Mutti. M,. and Simo, J.A.. 1994. Distribution, petrography and
geochemistry of early dolomite in eyc!ie shelf facies. "tales
Formation (Guadalupian). Capitan Reef Comple,i. USA, ln
Purser. B,. Tueker. M,. and Zenger. D. (eds,). Dolomiles A lo-
liime in Honorof Dotomieu. International Association of Sedimen-
tologtsts. Special Publication. 21. pp, 91 107,
Patterson. R.J,. and Kinsman. D,J,J,. 19K2, Formation of diagenetic
dolomite in coastal sabkha along the Arabian (Persian)
Gulf,
Atneriean
.Association
Petroleum Geologists Bulletin. 66:
28-43.
Purser, B,, Tucker. M,, and Zenger. D, (eds,). 1994, Dolomites—A
Volume in Honor if Dolomieu. Inlernational Association of
Sedimentologists. Special Publication. 21. 451pp.
Qing. H,, and Mountjoy. E,W,. 1992, Large-scale lluid (low in the
Mitldle Devonian Presqu'ile barrier. Western Canada Sedimentary
Basin, Geology. 20: 903 906,
Qing. H,. and Mountjoy. E,W,. 1994, Formation of coarsely crystal-
line,
hydrothermal dolomile reservoirs in the Presqu'ile Barrier.
Western Canada Seditnentary Basin. Amerieiin Assoeiaiion Peiro-
teum
Geologisis
Bullelin. 78: 55-17.
Radke. B,M,. and Mathis. R.L,, 1980. On ihe formation and
occurrenee of saddle dolomite. Journal of Sedimentary Petrology.
50:
1149 1168,
Schmoker. J,W,. and Halley. R,B.. 1982, Carbonate porosity versus
depth: a predictable relation for south Florida, Ameriean Assoeia-
iion Petroleum Geologists Bullelin. 66: 2561 2570.
Sanford. W,E,. and Konikov. L.F,. 1989, Porosity development in
coaslal carbonate aquifers. Geology. 17: 249 252.
Sanford. W,E,, Whitaker. F.A.. Smart. P.L.. and Jones. G,. 1998.
Numerical analysis of seawater circulation in carbonate ptalforms:
I. Geothenna! convection, Ameriean .lournalof Science. 298: 801-
828,
Shields. M,J.. and Brady. P,V,. 1995, Mass balance and tluid tlow
constraints on regional-scale dolomitizalion. Late Devonian.
Western Canada Sedimentary Basin, Bulletin of Canadian Petro-
leum Geology. Ai: 371 392,
Shukla. V.. and Baker. P,A. (eds,). 1988, Sedimeniology and Geoehem-
istry of Dolostones. Society of Fxonomic Paleontologists and
Mineralogists. Special Publication, 43. 266 pp,
Sibley. D,F.. and Gregg. J,M,. 1987, Classification of dolomile rock
textures. Journal of Seditiieniary Penology. 57: 967-975,
Spenecr-Cervalo. C. and Mullis. J,. 1992, Chemieal study of
tectonically controlled hydrothermal dolotnitization: an exatnple
froiti the Lessini Mountains. Italy,
Cieotoe,ische
Rundschau. 81/2:
347-370.
Stanley. S.M.. and Hardie. L.A.. 1999. Hypercalcification: paleontol-
ogy links plate tectonics and geochemistry to sedimeniology. GS.i
foday. 9(2): 2-7,
Stearns. N,D,. Slearns. H,T.. and Waring, G,A,. 1935. Flierma!
Springs
in
Ihe United.States. United Stales Geological Survey Waler
Supply Paper. 679-B. pp. ,59 191.
Smart. P,L,. Dauans. J,M,. and Whitaker. F,. 1988, Carbonate
dissolution in a modern mixing zone. Nature. 335: 811 813.
Swart. P,K.. and Melim. L.A.. 2000. The origin of dolotnites in
Tertiary sediments from the margin of Great Bahama Bank. Jour-
nal Sedimentary Research. 70: 738 748.
Usdowski. E,. 1994, Synthesis of dolomite—and geoehemical implica-
tions.
In Purser. B,. Tucker. M,. and Zenger. D, (cds,), Dolomiies-
.4 loliime in Honor of Dolomieu. International Association of
Sedimentolonists Special Publication 21. pp, 345-360,
Vahrenkamp. V,C,. Swart. P,K,. and Ruiz. J.. 1991, Episodic
dolomitization of Late Cenozoie carbonates in the Bahamas:
evidence from stronlium isotopes. Journalof Sedimeniary Petrology.
61:
1002-1014.
DUNES, EOLIAN 243
Vahrenkamp. V.C. and Swart. P.K.. 1994. Lale Ceno/oic dolomites of
the Bahamas: metastable analogttes Ibr the genesis of ancient plat-
form dolotnites. In Purser, B.. Tucker, M.. and Zenger. D, (eds,).
Dolomites .4
Volume
in Honor <f
Dolomieu.
International Associa-
tion of Sedimentologists. Special Publication. 2L pp. L13-153,
Vasconcelos. and McKenzie. J,A,. 1997. Microbial mediation of
modern dolomite precipitation and diagenesis under anoxie
conditions (Lagoa Vermelha. Rio tie Janeiro. Brazil), Journal of
Sedimentary Re.scanh. 67: 378 390.
Wanless. H,R,. 1979, Limestone response to stress: pressure solution
and dolomiti/ation, Joumahf Sedimentary
Petrology.
49: 437-462,
Ward. W,C.. and Halley. R.B,. 1985. Dolomitizalion in a mixing zone
of near-seawater eomposition. Late Pleistocene. Northeastern
Yucatan Peninsula. Jmirnaloj Sedimeniary Peirohgy. 55: 407-420.
Whitaker. F,F.. Smart. P.L,. Vahrenkamp. V,C,. Nieholson. H.. and
Wogelius. R.A,. 1994, Dolomitization by near-normal seawater,'
Field evidence from the Bahamas, In Purser. B.. Tucker. M,, and
Zenger. D. (eds.). Dolomites a \'olume in Honor of Dolomieu.
International Association of Seditnentologists. Speeial Publication.
21.
pp, 111-132,
White. D,E,. 1957, Thermal waters of volcanic origin, Geologiial
Society ot America Bullelin. 68: 1637-1658,
Xun. Z,, and Fairchild, I,J,, 1987, Mi,\ing zone dolomitization of
Devonian carbonates. Guangxi. South China, In Marshall. J,D,
(ed,),
Diagenesis in Sedimeniary Seijuenees. Geological Society of
London. Special Publication. 36. pp. 157-170,
Zenger. D.ll,. Dunham, J,B,. and Ethington. R,L, (eds,). 1980. Con-
cepts and Models of Dolomilizalion. Society Econotnic Paleontolo-
gists Mineralogists. Speeial Publieation. 28, 320 pp.
Dune types
Eoiian dtuics occur in five main morphologic types (Table Dl).
Althotigh sonic atilhors have argued lhat coastal duties have a
distinct tnorphology. many recent workers argue that the only
characteristic coastal dune form is the foredunc because il is an
integral parl of the complex ol' nearshore processes forrning
the beach-dune system (Bauer and Sherman. 1999). Three
varieties of each dune type can occur: simple (the basic form);
compound (superimposition of stnall dunes of the same type
on larger dunes): and comple,x (supcritnposilion of dilTerent
dune types on the primary fortn (e.g,. crescentic dunes on
linear dunes),
Eolian dunes occur in a self-organi/ed pattern (l-'igure 1336)
that develops as the response of sand surfaces to the u'ind
regime (especially its directional variability) and the supply of
sand (Werner and Kocurek. 1997). Relations between dune
types and wind regimes indicate that the main control of dune
type is the directional variability ot" the wind regime
(Figure D37), Grain size, vegetation cover, and sediment
supply play stibordinate roles,
N4igration oi'crescentic dnnes occurs as a result of erosion of
the stoss slope and deposition on the lee. Rates of dune
migration are inversely proportional to dtine height. Linear
Cross-references
Carbonate Mineralogy and Geochenii.stry
Cements and Cementation
Dedolomitization
Dolomite Textures
Evaporiies
Fitiid lncltisit>ns
I.sotopie Methods in Sedimentology
Porewaters in Sediments
Sabkha. Salt Flat. Salina
Seawater: Tetnporal Changes in the Major Solutes
Stains for Carbonate Minerals
DUNES, EOLIAN
Introduction
Eolian dunes form part ol a hierarchical system of bedforms
developed in wind-transported sand which comprises: (i) wind
ripples (spacing O.I
1
m): (ii) hidividual sitnple dunes or
superimposed dunes on cotnpound and comple,\ dunes
(spacing 50 5flOtn): and (iii) compoutid and eomplex dunes
(spacing >500tn). Dunes oeeur wherever there are sufficient
supplies of sand-sized seditnent, winds to transport that
sediment, and eonditions that promote deposition of the
transported sediment. These requirements are satisfied in two
main environtnents: (I) coastal areas with sandy beaches and
onshore winds: and (2) desert areas. Most dunes occur in
contiguous areas of eolian deposits called sand seas
(>10()km") or dune iields. Sand seas and dune (ields are
dynamic sedimentary bodies that form part of well-detined
regional- and local-scale seditnent transport systems in whieh
sand is tnoved by wind trom source areas {e,g,, distal fluvial
deposits, sandy beaches) via transport pathways to deposi-
tional sinks {Kocurek and Lancaster, 1999).
Table Dl A morpholof^ital 11 a ssi fixation ol" eolian sand dunes (atler
Lancaster
coarse
sand
Relative Sediment Thickness
CRESCENTIC
Barchans Cr^escentic Compound
ridges crescentic
dunes
LINEAR
Simple Compound Complex
-Straight Linear dunes Linear dunes
- SinuotJS
REVERSING DUNES •
STAR DUNES
ZIBARS
SAND
SHEETS
vegetation
anchored
NEBKHAS PARABOLIC
DUNES
source
bordering
topograpbic
influence
LUNETTES
FALLING
DUNES
CLIMBING
DUNES
244 DUNES, EOLIAN
Namib
Sand Sea
Frytierger
(1979)
X Star
Linear
Cre8centic
Figure D36 LANDSAT image of the soulhern pati ot" the Nainili
S^nd
Sea,
showing occurrence of different dune fypes as wind regime
changes from coast fo inland areas,
dunes do not migrate laterally, but extend downwind. Star
dunes are essentially stationary forms.
The simplest dune types and patterns form in areas
characterized by a narrow range of wind directions (unidirec-
tional wind regime). In the absence of vegetation, crescentic
dunes will be the dominant form. Parabolic dunes are the
equivalent in partially vegetated areas. Isolated creseentic
dunes or barchans occur in areas of litnited sand supply, and
coalesce laterally to form crescentic or barchanoid ridges that
consist ofa series of connected crescents in plan view as sand
supply increases. Larger forms with superimposed dimes are
termed compound crescentic dunes.
Linear dunes are characterized by their length (often tnorc
than 20km). sinuous crestline. parallelism, and regular
spacing, and high ratio of dune to interdune areas. They (brm
in areas of bitiiodal or wide unimodal wind regimes. The
essential meehanism for linear dune formation is the deflection
of winds that approach at an oblique angle to Ihe crest to flow
parallel to the lee side and transport sand along the dune. Thus
any winds from a 180" sector centered on the dune will be
diverted in this manner and cause the dune lo elongate
downwind (Tsoar. 1983),
Star dunes have a pyratnidal shape, with three or four
sinuous sharp-crested arms radiating from a central peak and
multiple avalanche faces. Star dunes oecur in multi-directional
or complex wind regimes and are the largest dunes in many
sand seas, reaching heights of more than 300 m. The
development of star dunes is strongly influenced by the high
degree of form—flow interaction which occurs as a result of
seasonal changes in wind direction, and the existence of a
major iee-side secondary circulation. Most ofthe erosion and
deposition involves the reworking of deposits deposited in the
previous wind season. Sand, once transported to the dune,
tends to slay there and add to its bulk (Lancaster. 1989).
Parabolic dunes are characterized by a U or V shape with a
"nose'"
of active sand and two partly vegetated arms that trail
star
X
X
•
•
Linear •
Ctescentic FtiOges
•
Barchan
3 50
cn
01
u
•5
30
T3
i 20
c
i? 10
to
CT 0
^ 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0.9
Directional variability of wind regime (RDP / DP)
Figure D37 Relations between dune types <ind wind regimes. From
VVasson and Hyde (19831 with Namib Sand Sea dafa superimposed
(symbols) and range of wind regimes lor three major dune types from
Fryberger (1979). Equivalent sand thickness is a measure ol the sand
available for dune building and represents the thickness of sand if the
dunes were levelled. The directional variability of the wind regime is
characterized by the ratio between resultant {RDP) and total sand drift
potential (DP),
up wind. They are common in rnany coastal dunefields and
semi arid inland areas and often develop from locahzed
blowouts in vegetated sand surfaces (Wolfe and David, 1997),
Other important dune types include nebkhas or hummock
dunes (common in may coastal dunefields) anchored by
vegetation; lunettes (often comprised of sand-sized clay pellets)
that form downwind of small playas; and a variety of
topographically controlled dunes (clitrtbing and falling dunes,
echo dunes).
Dune processes and dynamics
The development and tnorphology of all eolian dunes arc
determined by a complex series of interactions between dune
morphology, airflow, vegetation cover, and sediment transport
rates (Lancaster. 1995). As dunes grow, they project into the
atmospheric boundary layer so that streamlines are com-
pressed and winds accelerate up the stoss or windward slope.
The degree of flow acceleration (the speed-up factor) is
determined by the aspect ratio of the dune. Wind speed at
the crest of the dune is typically 1,1 times to 2,5 litnes that
measured itnmediately upwind of the dune (Figure D38(A)).
Flow acceleration coupled with effects of stream line curvature
on the windward slopes of dunes give rise lo an exponential
increase in sediment transport rates (Figure D38(B)) toward
the dune crest (McKenna Neuman etat., 1997). resulting in
erosion of the stoss slope. In the lee of dunes, there is a
complex pattern of flow separation, diversion, and re-aitach-
tnent determined by the angle between the wind and the dune
crest (angle of attack) and the dune aspect ratio (Sweet and
Kocurek, 1990). On high aspect ratio (steep) dunes, high
angles of attaek result in flow separation, and lower angles of
attack result in flow diversion. Low aspect ratio dunes are
characterized by flow expansion. Flow separation
(Figure D38(C)) results in the development of a back-flow