Originally published in Geophysical Corner, AAPG Explorer, June 2000.
Text and Figures, © Tectonic Analysis Ltd.
Part 1: Kinematic Analysis: The Next Step
James Pindell, Lorcan Kennan and Stephen Barrett.
For more information about Tectonic Analysis Ltd and our reports and services, please visit our website:
www.tectonicanalysis.com or email the authors: [email protected], [email protected]
An unfortunate fact of geology is that most datasets, including seismic, rarely allow for a unique
interpretation of a geological problem. Having to wrestle with multiple working hypotheses is perhaps
especially common in the structural arena, where one or another of many theoretical models of "ideal"
crustal deformation can be made to fit a given structural pattern. This can be frustrating and potentially
costly if the optimum exploration strategy is dependent upon the interpretation finally chosen.
Integration of multiple and diverse data sets is one popular approach to reducing the range of possible
interpretations, the goal being to minimize exploration risk. But on too many occasions, if your "best"
data set can't give you a clear solution, then mixing in diverse secondary data sets can muddle the
picture even further. Worse, this multifaceted picture may not be fully understood by anyone on the
work team, and the full implications of the "integrated solution," which will provide the basis of the
exploration model, might never be recognized.
It is widely recognized that broadening the scale of geological assessment to beyond the limits of the
block or field can help to constrain a unique solution to a given problem. Indeed, many plate tectonic
and structural processes evolve over scales far larger than most blocks, and to ignore the larger scale
can lead to serious misinterpretations. But broadening the scale of examination to beyond the block
remains, in many cases throughout industry, little more than a matter of describing what is out there. In
other words, mapping. As geologists, we all know that mapping is a key part of geology, but it is very
important to take the next step and understand how and why a given set of mapped structures
developed.
•
•
•
Can this help to resolve our interpretation of geological problems?
Can it tell us anything more about an exploration play?
Can it trigger the identification of new plays altogether?
We believe it can.
When we shift from trying to address the "what "questions of structural analysis into the "how"
questions, we move from static description into time-progressive kinematic analysis. Kinematic
analysis can be performed at all scales in geology - from mineral grains to tectonic plates - and it
embraces the motions of material undergoing geological change. Defining the motions of the plates and
crustal blocks, where possible, can tremendously facilitate understanding how certain types of
structures developed. Plate kinematics addresses the history of motion of the plates and blocks that
comprise or have comprised the earth's surface. Although plate kinematics is traditionally associated
with the oceans, it also can be applied successfully to areas of continental crust and margins of real
exploration interest.
1
In the late 1960s, one of the most exciting early realizations of the plate tectonic revolution was that the
ways in which plates move relative to each other, both past and present, are governed by a firm set of
predictive, or retrodictive, geometric rules. Plate kinematics gave us the power to quantitatively open
and close oceans, collide continents and evolve plate circuits in area-balanced models. Earth's
geological history became an intellectual playground for "plate pushers" who began to decipher Earth's
global tectonic evolution. However, all too often, these kinematic rules were either not applied,
misapplied or applied to inappropriate places, such that by 1980 many journal articles, no matter what
the discipline, ended with "bandwagon" Plate Tectonic Interpretation sections, which correctly came to
be viewed as mere arm waving. Similarly, industry decision-makers grew to be suspicious of such
tectonic scenarios - with good reason - and often ignored or discounted them. Thus, the potential of
kinematic analysis often was never reached. Sadly, these very powerful rules are no longer even taught
in many universities, and quantitative plate kinematic analysis is becoming something of a lost art.
Very powerful plate kinematic rules, however, do still exist. Here, in this first in a series of three
articles, we review some of these principles to provide the basis for exploring the power of kinematic
analysis.
In Figure 1a, we show a simple two-plate system in which block A moves NNE relative to B with time.
Displacement during the particular time interval of concern can be drawn as shown by the red vector
between the dots representing the plates. To palinspastically restore the offset back in time, we would
use the blue vector to retract the accrued measured offset. Progressing to a three-plate system, we must
consider the motions between the three pairs of plates. A simple analogy of this situation is to consider,
in Figure 1b, two runners, A and B, running from home plate to first and third base on a baseball
diamond. The displacement between home plate and runners A and B, respectively, is NE and NW, but
the relative motion between the two runners is east-west. A plate boundary separating plates
represented by the two runners would be extensional, with net E-W fault displacements.
In the three-plate example of Figure 1c, we can restore, moving back in time, two known offsets (A-C)
and (A-B) to determine the unknown offset between the third plate pair (B-C). The measured directions
and displacements of plates B and C are drawn relative to Plate A. Tieline B-C will then approximate
the net direction (NE) and displacement (76km) of the common B-C fault zone. If this happens to be a
thrust belt with the orientation as shown, then the strike-slip (blue, 30km) and convergent (red, 70km)
components of net motion can be inferred by construction of the right-triangle, thereby providing vital
information about overall structural style, with the expectation of dextral transpressive (combination of
strike-slip plus compression) strain partitioning at that thrust belt.
Finally, in the larger two-plate example of Figure 2, plates A and B diverge by seafloor spreading at the
ridge (red) and transcurrent motions at the transform faults (green). The continuations of the transforms
into adjacent oceanic crust are fracture zones where differential thermal subsidence occurs, but without
active strike-slip faulting. Ridge segments lie on great circles to the pole defining the plate separation,
whereas the transforms lie on small circles. The rate of plate separation and also of transcurrent
displacement at the transforms increases with distance from the pole. Transforms also become
straighter as distance increases from the pole of rotation.
Subsequent articles in this series will apply these principles to two well-known oil provinces,
Colombia/western Venezuela and the Gulf of Mexico, showing how formal kinematic analysis can
offer some of the most sound constraints available to guide and to favor certain geological
interpretations over others. Further, it can provide the basis for defining or rejecting play concepts,
therefore strongly influencing exploration strategy.
2
www.tectonicanalysis.com
A
Accrued
Offset:
2-PLATE SYSTEM
A
B
A
B
A
B
Restored
Offset
B
Second
Base
Third
Base
First
Base
RUNNER
B
RUNNER
A
Home
C
70km
3-PLATE SYSTEM
A (?)
30km
B (70)
C (40)
B
76km
C
70km
40km
A
100 km
Figure 1. Examples of vector displacement diagrams
for two and three-plate systems.
Published in AAPG Explorer, v. 21, June 2000
www.tectonicanalysis.com
Concentric Small
Circles about Pole
Ridge
Segments
Pole of Rotation
Transforms
Fracture
Zones
PLATE A
PLATE B
Great Circles
through Pole
Figure 2. Relationships between pole of rotation,
great circles, ridge segments, small circles,
transforms, and fracture zones in a two-plate system.
Published in AAPG Explorer, v. 21, June 2000
Originally published in Geophysical Corner, AAPG Explorer, July 2000.
Text and Figures, © Tectonic Analysis Ltd.
Part 2: Kinematics a Key To Unlocking Plays
James Pindell, Lorcan Kennan and Stephen Barrett.
For more information about Tectonic Analysis Ltd and our reports and services, please visit our website:
www.tectonicanalysis.com or email the authors: [email protected], [email protected]
Last month's Geophysical Corner outlined some of the principles and methods of kinematic analysis as
a means of better deciphering the structural history of basins. In this second article, we apply some of
the principles of kinematic analysis to the first of our example areas: the northern Andes of Colombia
and western Venezuela. We also will illustrate some uses and benefits of this analysis to petroleum
geology and exploration in continental settings. When applied to continental areas, kinematic analysis
provides map-view palinspastic reconstructions of deformed regions prior to the deformation(s),
analogous to balancing cross sections in the vertical plane. Two very useful applications of continental
block kinematics for exploration are:
•
•
To allow more accurate plotting of paleofacies for times prior to deformations.
To allow more rigorous reassembly of continental blocks that have become separated during rifting,
enhancing the understanding of the development of hydrocarbon-bearing continental margins.
Here we show a set of simple steps for restoring the northern Andean ranges and basins for Early
Oligocene and earlier time, prior to the majority of "Andean" deformation. Note that variations in the
reconstruction will derive from applying different numbers of steps (accuracy can be increased by
accounting for more fault motions between more blocks), and also from adjusting various input
parameters, such as magnitudes of strike-slip offset on certain fault zones. A reference frame is needed
to begin: In this example, Andean motions are assessed relative to the Guyana Shield.
First, we address the relative motion of the Maracaibo Block by assessing displacement in the Mérida
Andes. Figure 1 shows the ca. 150 km dextral offset across the Mérida Andes of the "Eocene
thrustbelt,". In addition, shortening in the Mérida Andes has been estimated as about 40 km. Thus, in
the Early Oligocene, the Maracaibo Block lay significantly farther southwest relative to the Shield than
it does today. In figure 2, we construct a tie line between the Shield and Maracaibo Block by
performing vector addition of the strike-slip (150 km) and thrust (40 km) components. Because we
wish to restore the accrued offset (155 km), we draw the tie lines opposite to the real-life sense of fault
displacements, i.e., moving back in time.
Having defined the Oligocene paleoposition of Maracaibo, our next concern is the Perijá Range,
deformation of which accounts for movements between the Maracaibo Block and the Santa Marta
Massif. Estimates of post-Early Oligocene shortening are ca. 25 km, as shown by the Perijá vector in
figure 2. Thus, displacing Santa Marta Massif to the west-northwest of Maracaibo by 25 km gives the
Early Oligocene position of Santa Marta relative to both Maracaibo and the Shield. Next, the Santa
Marta strike-slip fault displaces the Santa Marta Massif Block from the northern part of Colombia's
Central Cordillera. Left-lateral offset of about 110 km (figures 1,2) is believed to have occurred on this
fault zone since Late Oligocene. This strain is transferred into the Eastern Cordillera along the south3
southeast continuation of the fault - the Bucaramanga Fault. Interestingly, this fault is flanked by the
high, compressive topography of Santander Massif - because the Bucaramanga Fault defines the
boundary between the Central Cordillera and the Maracaibo Block, not the Santa Marta Block. For
simplicity in figure 2, the trend shown for the fault (in orange) defines only the total strain between
those blocks, i.e. the sum of the strike-slip and orthogonal components of relative motion. Finally, we
restore Colombia's Guajira Block, also relative to Santa Marta, by removing about 125 km of dextral
shear on the Oca Fault to realign the western edges basement in the two blocks prior to faulting.
With just these simple considerations, and assuming that only minor vertical axis rotation of these
blocks has occurred during their relative motions, we can now fill out other tie lines in the vector "nest"
of figure 2 to define offsets between other pairs of blocks in the system. For example, the total strain in
the Eastern Cordillera since the Oligocene is seen to be ca. 200 km toward the east-southeast (red tie
line). This can be broken down into components of orthogonal and strike-parallel strain of 180 km
(blue line) and 100 km (green line), which translates geologically into shortening (180 km) and dextral
shear (100 km). This value of shortening (180 km) falls in the middle of the range of published values
of estimated shortening in Eastern Cordillera. Thus, vector nests such as figure 2 can be used to help
choose between alternative balanced cross section models assessing shortening. In addition, it also
allows detection and estimation of the strike-slip component, which usually cannot be seen in cross
sections. Our inferred dextral shear in the Eastern Cordillera is supported by seismicity, GPS data and
field observations.
A pre-Andean (i.e., pre-Late Oligocene) palinspastic reconstruction of the northern Andes continental
region (figure 3) now can be made by restoring the motions of the blocks defined in figure 2. The
known limit of pre-Mesozoic continental crust has been identified in figure 3 to show the pre-Andean
geometry of the northern Andes "autochthon," to which a number of oceanic terranes have been
accreted in the Cenozoic. Additional information can now be added to better focus the picture. We can,
for example, draw the occurrence of Eocene formations, sedimentary facies and paleoenvironments on
our reconstruction in order to build palinspastically accurate models of regional Eocene depositional
systems. This practice also allows better sequence stratigraphic interpretation and correlation at the
regional scale, which is helpful to determining migration pathways through the strata. Also, the
depositional models can be compared more meaningfully to modern analogues and analyzed for
implications concerning reservoir potential, such as sand body orientation, sinuosity, flow direction,
sand grain provenance and sediment maturity. Finally, the reconstruction also allows a better
interpretation of Cretaceous source rock character, quality and original areal extent.
Using the same block/plate restoration technique, we can depict Eocene-aged structures and the Eocene
position of the Caribbean Plate relative to South America, to better understand the driving forces of
Eocene sedimentation patterns and deformation. Figure 4 thus shows the Caribbean Plate driving an
Eocene foredeep basin in the northern Maracaibo area - much like today's Persian Gulf - which caused
an important early hydrocarbon maturation event in western Venezuela and Colombia's Cesar Basin.
Figure 4 also shows depositional systems with important reservoir facies belts at the Middle to Late
Eocene boundary, as well as the existence, continuity and origin of an Eocene "Maracaibo Tar Belt" in
western Venezuela (also recognized in Middle to Late Eocene field sections). The concept of this
"textbook" foredeep basin for the Eocene of Maracaibo Basin had remained darkly veiled for decades
by today's grossly different geography.
Next month, we will use plate kinematics to reconstruct Africa and South America, and to
progressively close the Atlantic Ocean during Mesozoic times, in order to set the stage for tracing the
evolution of the Gulf of Mexico and the Florida/Bahamas region in our fourth article of the series.
4
www.tectonicanalysis.com
-77
-76
-75
-74
-73
-72
-71
-70
-69
13
13
Guajira
12
12
Oca Fault
11
Santa
Marta
11
Falcón
Basin
Perijá
Andes
Santa
lts
oF
ar
ap
8
,C
Ca. 150 km
offset of
Caribbean
Nappes
and Eocene
facies belts
7
Bo
ela bia
zu
ne olom
C
co
Ve
Eastern
Cordillera
Terranes to
the west of
the Romeral
Fault are
allochthonous
5
nó
ault
Central
Cordillera
6
Merida
Andes
nga F
7
9
rama
8
Maracaibo
-Buca
Romeral Fault
Marta
9
10
au
10
6
5
Guyana
Shield
lt
au
eF
gu
4
Iba
4
3
3
-77
-76
-75
-74
-73
-72
-71
-70
-69
Figure 1. Map of northern South America showing main crustal
blocks, separated by lithospheric fault zones, under relative
motion during Late Oligocene to Recent Andean Orogeny.
Published in AAPG Explorer, v. 21, July 2000
www.tectonicanalysis.com
100 km
Partitioned components,
Eastern "Andean" strain
Northern
Central
Cordillera
110 km (Santa
Marta Ft.)
180 km
200 km
(Eastern Cordillera)
120 km
155 km (Merida)
(Bucaramanga Ft.)
Santa
Marta
25 km (Perij‡)
Maracaibo
Guyana
Shield
150 km
40 km
Km
0
50
100
150
Figure 2. Vector "nest" restoring displacements of northern
Andean blocks along faults during Andean orogenesis. Heavy
dots denote blocks, tie lines restore net azimuth and magnitude of fault displacements, moving back in time. Mérida
Andes and Eastern Cordillera deformation is shown partitioned into strike-orthogonal and strike-parallel components.
Published in AAPG Explorer, v. 21, July 2000
www.tectonicanalysis.com
Palinspastic Grid
Northern Andes, 25 Ma
12°
Northern edge
of autochthon
Santa
Marta
11°
Valledupar
Maracaibo
10°
Paleo-positions of
Lake Maracaibo
and border
9°
7°
Bucaramanga
8°
Medell’n
6°
Honda
Western edge
of autochthon
Cusiana field
Fault
Bogot‡
Ca–o
Lim—n
field
5°
4°
C a l i
3°
Rom
era
l
Paleo-positions of
present courses of
Rios Cauca and
Magdalena
2°
1°
Orito field
76°
77°
72°
75°
74°
73°
0°
Figure 3. Oligocene reconstruction of pre-Mesozoic continental basement, northern Andes, based on vector displacements
from Figure 2. "Retro-deformed" grid of longitude and latitude lines is created by smoothing the lines across block boundaries after block restoration. Red outline is present day South America for comparison. Cities, fields, rivers and geographic
features (blue) are shown in palinspastic coordinates to help show Oligocene paleogeography.
Published in AAPG Explorer, v. 21, July 2000
www.tectonicanalysis.com
14.0°
Caribbean
Plate
Antilles
Arc
Juvenile
Barbados
Prism
?
70
Caribbean
Nappes
ar
Misoa-Pauji
Foreland Basin
ac
aib
o
Tar Belt
?
Flexural
Bulge
Guyana
Shield
After Pindell et al., 1998. SEPM Special Paper 58
6.0°
-74.0°
Serrania and
Trinidad
reentrants
Caracas
10.0°
M
ProtoCaribbean
Seaway
-70.0°
-66.0°
ENVIRONMENT
Emergent
Alluvial plain
Coastal fringe
Sandy shallow sea
Muddy outer shelf
Deep water
Carbonates
-62.0°
Figure 4. Paleogeographic map of western Venezuela and northern
Colombia, showing the position of the Caribbean Plate and main
depositional units during Eocene time. Note the similarity with
today’s "Persian Gulf".
Published in AAPG Explorer, v. 21, July 2000
Originally published in Geophysical Corner, AAPG Explorer, July 2000.
Text and Figures, © Tectonic Analysis Ltd.
Part 3: A Removal-Restoration Project
James Pindell, Lorcan Kennan and Stephen Barrett.
For more information about Tectonic Analysis Ltd and our reports and services, please visit our website:
www.tectonicanalysis.com or email the authors: [email protected], [email protected]
In our first two articles in this series we showed how vector triangles and rotation poles can be used to
constrain the motions of continental blocks and plates, and we reconstructed the pre-Andean
(Oligocene) shape of northern South America. This month we show the importance of removing postrift sedimentary sections and restoring crustal extension when approximating the pre-rift shapes of
continental blocks and margins.
First we'll show how this can be done in a simple way, and then we'll apply the method to a rifted
margin pair - the equatorial margins of Africa and South America - to derive a pre-Aptian
reconstruction of the northern parts of those two continents. Prior to the equatorial Atlantic break-up
during the Aptian, the northern parts of these two continents were essentially a single block. We can
use the Euler rotation poles defined by marine magnetic anomalies and fracture zones in the central
North Atlantic to rotate the reconstructed shape of Africa/South America back toward North America.
This process, when combined with the pre-Andean palinspastic reconstruction of the northern Andes
from last month's article, provides a quantitative kinematic framework in which to base models for the
Mesozoic evolution of the Gulf of Mexico, Mexico and nuclear Central America, the Florida/Bahamas
region, the Proto-Caribbean Seaway and northern South America. Continental rifting reflects
divergence of relatively stable portions of crust. This is accommodated by crustal extension at shallow
levels (typically less than 15 km), by normal faulting and at depth by ductile stretching of the lower
crust and upper mantle. The end result is lithospheric thinning at the rift; we usually see overall tectonic
subsidence of the surface, elevation of the asthenosphere, increased heat flow and, sometimes,
volcanism. At the surface, fault-bounded grabens initially fill with red beds, if subaerial, as rifting
proceeds. These are then overlapped by "thermal sag" sedimentary sections driven largely by cooling
of the asthenosphere, plus the loading effect of the sediments themselves. Where extension is
sufficiently large, oceanic crust is created and the two portions of continental crust drift apart. Where
rifting does not reach this stage, we are left with intra-continental basins. Sediment thickness at the
rifted margins that flank ocean basins can exceed 16 km.
If sediment supply is sufficient - for instance, near deltas or adjacent to high-relief topography in wet
climates - the position of passive margin features such as the shelf-slope break can change significantly
with time, growing out from the coast and well beyond the original limits of the continental crust
(figure 1a). Although used for Bullard's famous reconstruction of the Atlantic margins (1965), this is
why it is not satisfactory in quantitative kinematic analysis to merely realign a given bathymetric
contour along opposed pairs of passive margins. To gain a much closer approximation of the shapes of
rifted margins to fit together for a more precise pre-rift geometry, we must construct cross sections of
rifted margins that depict the thicknesses of the water column, the sedimentary section, and the crust.
Water depth and total sedimentary section are often known from geophysical studies at passive
margins. The position of the Moho (base of the crust) can be crudely estimated by the balancing of
5
water, sediment, crust and mantle using Airy isostatic calculations (figures 1b,c) and, where gravity
data or detailed sedimentological data are available, refined by taking into account crustal flexure and
sediment compaction. Once the cross-sectional shape of the rifted margin's crust is inferred, the syn-rift
extension in basement can be removed by restoring the cross-sectional area of the rifted margin
shoulder back to an unstretched beam of continental crust. Again, a crude calculation can assume this
started at or near sea-level, and more refined calculations could take account of surface elevation, water
depth prior to rifting and variations in initial crustal thickness or density. This identifies the position
within that cross section that defines the pre-rift edge of the continental block. When plotted at several
points along a particular margin, we can estimate the pre-rift shape of the continental margins. This can
then be rotated towards the opposing margin using plate kinematic methods to show pre-rift geological
relationships - and to provide a starting point for modeling the ensuing basin evolution.
Figure 2 shows the net result of this method when applied to the rifted margins of the Equatorial
Atlantic. The method is particularly important along the shelves at the mouths of the Niger and
Amazon rivers, where the sedimentary thickness exceeds 10 km over large areas. Note that the ParaMaranhao Platform is a piece of the West African Craton stranded on South America as the Equatorial
Atlantic opened. A satisfactory fit can be achieved to an accuracy of perhaps 50 km. For comparison,
the inset of figure 2 shows the classic Bullard reconstruction of the two continents, with the pre-rift
shapes of basement shown rather than the 2,000-meter isobath employed by Bullard. The inferred
underfit in the Bullard reconstruction approaches 500 km. Because continental reassembly in the Gulf
of Mexico region is achieved by rotating the Africa-South America reconstruction back toward North
America using Central Atlantic kinematic data, the difference between the two approaches will affect
the final reassembly as profoundly as any other kinematic parameter. Marine magnetic anomalies and
fracture zone traces are used in the oceans to track the past velocity and flowpath, respectively, of pairs
of plates separated by seafloor spreading.
Figure 3 shows a series of reconstructions of our united Africa-South America supercontinent and
North America for Aptian and older times, prior to Equatorial Atlantic break up. Some of the positions
are interpolated or extrapolated from the marine data to provide key time slices such as Triassic
Pangean continental closure, and late Callovian/Early Oxfordian salt deposition in the Gulf. The
analysis tells us how fast and in what direction the continents separated, which in turn constrains the
geometry of ridge systems between the Americas, and also the size and shape of the inter-American
gap through time. Finally, also shown on figure 3 is the pre-rift palinspastic shape of the northern
Andes region superimposed on South America for the Late Triassic time slice. This was drawn by
taking last month's reconstruction (i.e. prior to Cenozoic shortening and strike-slip) and modifying it
for pre-rift time by applying the methodology of figure 1 (assuming an ENE-WSW extension
direction). The relationship of North and South America at this time is important, because it defines a
line separating two parts of Mexico. The part of Mexico overlapped during Late Triassic time by South
America must have migrated into today's position as a function of Gulf of Mexico evolution,
Cordilleran terrane migration, and/or Sierra Madre/Chiapas shortening history. Parts of Mexico not
overlapped by South America during the Triassic may have been in place relative to today's geography,
but were not necessarily so. From figure 3, the fact that the formation of the Gulf of Mexico was
completed by early Cretaceous time implies that Jurassic plate boundary systems active in the Gulf
until then probably also controlled many primary elements of the evolution of Mexico.
Thus, the stage is set for us next month to use the kinematic constraints developed here to reconstruct
western Pangea and to trace the Mesozoic plate-kinematic evolution of the Gulf of Mexico, eastern
Mexico, the Florida/Bahamas region and the Proto-Caribbean Seaway in our final article of the series.
6
www.tectonicanalysis.com
A
Note misfit if we matched Shelf-Slope Break instead
of Pre-rift reconstructed continental Edge
Misfit
Shelf-Slope Break
Pre-stretch Edge Continent-Ocean Boundary
of Continent
B
Post-rift Crustal Profile
Water/sediment
Water/sediment
Stretched
Continental
Crust
Unstretched
Mantle
Stretched
Continental
Crust
Ocean
Crust
ContinentOcean
Boundary
C
Required
Restoration
Complete removal of all stretching
Reconstructed
continental edge
Figure 1. A, Cartoon section showing how passive margin sediments (deltas,
turbidites, carbonate banks) can prograde far beyond the original position of the continental edge. B and C, Simple method of estimating and restoring crustal extension
during rifting and passive margin formation. The cross-sectional area of the stretched
crust (hatch-pattern) must equal that of the unstretched crust after sediment, water and
mantle have been removed from the cross-section.
Published in AAPG Explorer, v. 21, August 2000
www.tectonicanalysis.com
10
5
0
5 After Pindell 1985, 10
Tectonics v4, p1
West African Craton
AFRICA
Central
Atlantic
Oceanic
Crust
10
Restored crustal
limits, reconstructed
prior to rifting
River
Niger
5
Rotational restoration of
Para-Maranhaõ Platform
(originally part of W. African
Craton) to close Marajo
Basin extension and avoid
overlap with Ivory Coast
AFRICA
Inset:
Mismatch (underfit) in
Bullard fit when crustal
stretching and sediment
are removed
500 km
River
Amazon
Present
Coastlines
NORTHEAST
SOUTH AMERICA
500 km
Bullard (1965)
fit coastlines
SOUTH
AMERICA
55
50 10
45
15 40
Figure 2. Pre-Aptian Equatorial Atlantic reconstruction in which the restored pre-rift limits
of continental crust (ie, methodology of Figure 1) are juxtaposed. Note resulting simple
geometry for Aptian rifting. Inset: Bullard (1965) reconstruction of the two continents,
which realigned the 2,000 m isobaths of today’s passive margins (not shown), showing the
pre-rift limits of continental crust for each, as well as the large region of continental underfit
in the absence of the sedimentary sections.
Published in AAPG Explorer, v. 21, August 2000
Published in AAPG Explorer, v. 21, August 2000
M16 Anomaly (140 Ma)
M10 Anomaly (130 Ma)
M0 Anomaly (118 Ma)
Present Day Coastline
Continent-Ocean Boundary (?200 Ma)
Blake Spur Anomaly (?165 Ma)
Early Oxfordian, interpolated (160 Ma)
M25 Anomaly (156 Ma)
Based on data presented in Pindell et al. 1988 (Tectonophysics v.155 p.121)
M21 Anomaly (149 Ma)
KEY TO RECONSTRUCTION AGES
?
Late Triassic, total closure (250 Ma)
Pre-rift shape of
northern South
America at time
of total closure
Central American
terranes to SE of
this line must be
allochthonous
Pre-rift shape
of Gulf Coast
80
500
1000
1500
2000
2500
3000
160
TIME (MA)
120
200
M21
M25
BS
COB
M0
M10
M16
SEPARATION DISTANCE OVER
TIME BETWEEN THE AMERICAS
(SLOPE = VELOCITY)
Present Day Coastline
Unless we restore the south Florida
offshore back to the NW we are still
left with a gap in the eastern Gulf
DISTANCE (KM)
Yucatan
fits here
www.tectonicanalysis.com
Figure 3. Successive pre-Aptian reconstructions of Gondwana and North America,
using the Equatorial Atlantic fit of Figure 2. This analysis provides a quantitative
framework in which to build more locally detailed models of the evolution of the Gulf
of Mexico and surrounding areas. Note pre-Andean/pre-rift restoration of the northern
Andes on the Triassic position of South America: this defines how much of Mexico is
definitely allochthonous versus how much is potentially , but not necessarily,
autochthonous.
Originally published in Geophysical Corner, AAPG Explorer, July 2000.
Text and Figures, © Tectonic Analysis Ltd.
Part 4: Putting It All Together Again
James Pindell, Lorcan Kennan and Stephen Barrett.
For more information about Tectonic Analysis Ltd and our reports and services, please visit our website:
www.tectonicanalysis.com or email the authors: [email protected], [email protected]
In the second and third parts of this series we defined a kinematic framework for the evolution of the
Gulf of Mexico region by restoring Andean deformations and progressively closing the Atlantic Ocean.
This month, we further evolve this framework to build a palinspastically quantitative reassembly of
continents and continental blocks that were separated during the Mesozoic rifting and subsequent drift
in the Gulf of Mexico region - key features of which are shown in figure 1.
Figures 2-4 show primary developmental stages in the Gulf's evolution:
• Post-Gulf formation (figure 2).
• Post-salt/pre-seafloor spreading (figure 3).
• Early syn-rift (figure 4).
The kinematic elements applicable to the reconstructions are as follows.
First, our Oligocene reconstruction of northern South America (article two) is modified for Late
Jurassic and Cretaceous time by removing island arc and other terranes that accreted to in the Late
Cretaceous and Early Tertiary (shape shown in figures 2 and 3). We then estimate and restore Jurassic
extension in rift basins of the Andes (using principles outlined in the August EXPLORER, which gives
us an Early Jurassic shape for the northern Andes that can be closed against North America (figure 4).
Second, figures 2-4 show that Florida, the Blake Plateau and the Bahamas (and the "Cuban autochthon"
beneath the Cuban arc) were strongly controlled by fracture zone trends of the early Atlantic. Here,
plate separation was achieved by NW-SE stretching of crustal blocks separated by transcurrent faults.
Middle Jurassic basalt extrusion was commonplace in zones of high stretching. Each crustal "corridor"
between transcurrent faults underwent different amounts of stretching and displacements relative to the
others. The conjugate margin to the Southern Bahamas flank is the transcurrent margin of Guyana.
Third, unlike the Florida region, the Yucatan Block moved independently of the larger continents - in
two distinct stages - as the Gulf opened. At the time of figure 4, there is only a small range of paleopositions in which Yucatan can fit without overlap of palinspastically restored (i.e., rift-related
stretching removed) areas of continental crusts. This position can be achieved by rotating present-day
Yucatan clockwise about "pole A" (figure 4), which closes most of the Gulf by placing Yucatan snugly
against the northeast Mexico-Texas-northwest Florida paleo-margin. It definitely does not, however,
close the southeastern Gulf. There, the crust of South Florida - including that of the "Tampa Arch" must be retracted northwestward against Yucatan and out of an overlap position with Demerara Rise,
off the Guyana margin. Thus, the southernmost crustal corridor of the Bahamas must have migrated
SE, probably along our "Everglades Fracture Zone" (figure 1) between the times of figures 3 and 4.
7
Fourth, the geology of eastern Mexico and the occurrence of Louann and Campeche salt suggest that
the Gulf opened in two stages. The first, or syn-rift, stage - between the times of figures 3 and 4involved intra-continental stretching between Yucatan and North America about "pole B1," and
between Yucatan and South America about "pole B2," (figure 4). This migration defined an arcuate
transcurrent trend of basement contours along the northern Tamaulipas Arch in south Texas. Also
sinistral shear in the Louisiana-Mississippi area, allowed for minor counterclockwise rotation of the
Wiggins and Middle Grounds arches (figures 1 and 4) and the associated formation of the wedge
shaped East Mississippi and Apalachicola salt basins to the north of each, respectively. This syn-rift
stage about "pole B1" can be modeled satisfactorily to Early Oxfordian time to achieve a good
reconstruction of the Louann and Campeche salt provinces flanking the central Gulf (figures 1 and 3).
There is no need to invoke significant salt deposition on ocean crust in the Gulf. Also, during this stage,
the southern Bahamas crustal corridor migrated southeast while undergoing internal stretching – with
the Everglades fracture zone and Guyana marginal fault zone both active at this time. The migration of
Yucatan to its present position requires that eastern Mexico was a transform rather than a rifted margin.
Also, Yucatan did not have the Chiapas Massif attached to it during the syn-rift phase. Why?
First, we cannot satisfactorily fit a combinedYucatan/Chiapas Massif into the northern Gulf, especially
when we reverse the effect of Cenozoic shortening in Sierra de Chiapas. Second, we believe that the
Chiapas syn-rift salt basin is best explained by early transtension along a crustal scale fault beneath it.
The second stage of Yucatan motion began about "pole C" of figure 3, in the Early Oxfordian, at the
end of salt deposition. This second stage of motion and its pole of rotation are constrained by:
•
•
•
Geophysical data along the eastern Mexican margin, which show an abrupt NNW-SSE trending
ocean-continent boundary.
Magnetic anomaly data in the eastern Gulf.
Displacement of the once-adjacent margins of the Louann and Campeche salt basins.
The Chiapas Massif was picked up by Yucatan in this stage as a consequence of the onset of seafloor
spreading in the Central Gulf, and because the pole of rotation changed in Stage 2, the orientation and
position of transforms also changed. This new phase of motion had a more southerly direction than the
previous one. The spreading ridge almost reached the Mexican coast and, hence, the new transform
along eastern Mexico picked up an additional wedge of crust - Chiapas Massif - which had been
emplaced there during the syn-rift phase by sinistral transcurrent motions within greater Mexico.
As with the Gulf of Mexico, the creation of the "Proto-Caribbean Basin" also must have involved a
rotational opening between Yucatan and Venezuela-Trinidad. In figures 2-4, we show the approximate
flowlines along which this basin opened, and hypothetical Jurassic rifted margin geometry now wholly
overthrust by Caribbean terranes. Many elements of northern South America's and eastern Yucatan's
hydrocarbon potential pertain directly to the geometries of these rifted margins, such as the positions of
marginal re-entrants that define differing stratigraphic sequences due to differing subsidence histories.
Our working Gulf kinematic model has some interesting implications for exploration.
First, the Eastern Mexican margin (unlike that of Texas) was a Jurassic fracture zone in the north
(Burgos-Tampico basins) and a transform - with active structuring until its Early Cretaceous death - in
the south (Veracruz Basin). Heat flow, subsidence history, occurrence of salt, distribution/thickness of
Late Jurassic source rocks and basement controls on future structural development will all vary along
8
strike along this margin due to differing crustal properties and histories. In the U.S. Gulf margins, early
syn-rift stretching was NNW-SSE until Early Oxfordian times, but most of the stretching toward the
end of this phase occurred well offshore.
Second, although salt deposition is generally assumed to be of Callovian age, there is little evidence of
open marine conditions in the Gulf margins until upper Oxfordian (Norphlet-Smackover transition),
and thus salt deposition may have continued until Early Oxfordian. Our Early Oxfordian reconstruction
accommodates known salt occurrence in the Gulf ("salt fit"); hence, onset of seafloor spreading, the
change in the Yucatan-North America pole position, separation of Louann and Campeche salt
provinces, and initiation of open marine conditions were nearly coeval and possibly causally related.
Third, although the syn-rift stretching of the Florida Shelf region was NW-SE, the extension direction
in the deep eastern Gulf during stage 2 (seafloor spreading) was NE-SW about a nearby pole, such that
small circles (transform traces) should be arcuate and convex to the northwest. In the Jurassic, the
southern Bahamian margin (beneath Cuban overthrust terranes) experienced sinistral strike-slip
tectonics along the Guyana margin of South America, followed by the eastward migration of a Late
Jurassic seafloor spreading ridge (Yucatan/South America boundary) along the western half of the
overthrust zone. The transform nature of this Jurassic margin should be considered in interpretations of
the Paleogene development of the Cuban thrust belt, Mesozoic source rock paleogeography and oil
migration pathways during Eocene maturation.
In the Proto-Caribbean, the kinematics require westward-propagating Early and Middle Jurassic rifting,
followed by Late Jurassic seafloor spreading. The trends of marginal re-entrants such as that defined by
the Urica basement transfer zone are defined by the first stage of Yucatan's motion. Further,
Venezuela-Trinidad's passive margin section is predicted to have existed from the end of Middle
Jurassic, not Cretaceous as is commonly thought. A several kilometer-thick, probable Late Jurassic
shelf section in Eastern Venezuela has not received much attention from exploration, and the
"Berriasian or older" salt in Gulf of Paria could be Middle Jurassic (as is the salt in the Bahamas,
Guinea Plateau and Demerara Rise and Tacatú Basin). Note the proximity of these areas on figure 4. In
Sierra Guaniguanico of western Cuba, the conjugate margin of Eastern Venezuela, the lower Middle
Jurassic San Cayetano strata indicate the existence of a juvenile passive margin of that age, becoming
fully marine for Late Jurassic, as predicted here for Venezuela and Trinidad.
In summary, regional plate kinematic analysis is extremely cost-effective and deserves an important
role in the exploration of complex areas, both early on and long-term. The kinds of implications we
have drawn here also can be made from kinematic analysis in other parts of the world. When applied
properly to appropriate areas, it is not arm waving.
Much can be gleaned about:
• Fault styles and displacements.
• Basement types and associated parameters such as early heat flow.
• Systematics of regional reservoir-bearing depositional patterns.
• The relative ages of classes of structures, etc.
And all that is gleaned can lead to the creation or dismissal of numerous play concepts. In addition, an
explorationist with a comprehensive kinematic framework available to him or her will work more
confidently - and therefore, more efficiently - on nearly all other aspects of the exploration process.
Finally, in frontier evaluation programs, regional kinematic analysis may not tell you exactly where to
drill, but it can often help to tell you where not to drill.
9
www.tectonicanalysis.com
Limit of Salt
Monroe Uplift
Sabine Uplift
Apalachicola
Bsn
Mississippi Bsn
Llanos High
Lr.
K.
eef
R
Middle
Grounds
Arch
Wiggins Arch
Louann
Salt basin
)
kso
nvi
lle
PC
FZ
Pz
15
PC
Pz
Sigsbee
Mz
PC
Pz
Tamaulipas
Arch
Golden
Lane
Fra Blak
ctu e S
re Z pu
one r
(FZ
Jac
Magnetic Edge
of Ocean Crust
Burgos
Basin
Limit of
basement
data
20'
FZ
Relay (pull-apart)
Tampa Arch
East
Mexican
FZ
"S.
Limit of Salt
Yucatán Block
Campeche
Salt basin
S. Guaniguanico
(Yucatán Block)
Cu
ban
FZ
"
Veracruz Bsn
Chiapas Foldbelt
Isth. Tehuentepec
Chiapas Massif
Figure 1. Present day map of Gulf of Mexico region, showing key geological
°
elements addressed in text. Note abrupt terminations of known basement
units in southern Florida. Also note change in trend of East Mexican Marginal Fault Zone supporting the concept of two stages of Gulf evolution; basement structure contour data preclude any E-W faults in Mexico from entering
the Gulf during the sea-floor spreading stage. Digital bathymetry/relief after
Sandwell and Smith (1997), other features from multiple sources.
Published in AAPG Explorer, v. 21, September 2000
Important Note: We have modified Figure 1 since original submission for publication in
the Explorer. The reason is that the magnetic and gravity patterns of the Sarasota
Arch, off southwest Florida, do not appear to be offset or interrupted along the projected trace of our "Everglades Fracture Zone". Thus, it may be preferable to consider that SE-ward motion of the Florida Straits block occurred by sinistral shear jumping
northward at the Northeast Basin, which would make that basin a pull-apart basin for
Middle Jurassic time, prior to seafloor spreading in the central Gulf of Mexico.
www.tectonicanalysis.com
40
105
100
95
90
85
80
75
Early Cretaceous 130 Ma
North America
(present day coordinates)
35
Unstretched
Stretched continent
Ocean crust
Basalt Plateau
Thin Salt
Thick Salt
70
40
35
Neogene halokinesis (Sigsbee)
Chihuahua,
Sabinas
Basins
Magnetic
Anomaly M10
30
30
Yucatán-NOAM pole
(Mid. Jur. - E. Cret.)
Yucatán
20
Mexican
back-arc
Proto-Caribbean
Seaway
20
Transitional?
Ch
iap
as
Chiapas Massif
in final position
15
Reef
edge
Chortis
10
105
New ridge is cut
once rotation stops
Jamaica, Cuba
100
15
South America
Simplified from Pindell and Kennan, 2000, in prep.
95
90
85
80
75
10
70
Figure 2. Early Cretaceous (Valanginian) reconstruction of the Gulf
of Mexico and Proto-Caribbean region. Post-Gulf formation stage,
when seafloor spreading in the Gulf had ceased but was continuing
in the Proto-Caribbean seaway.
Published in AAPG Explorer, v. 21, October 2000
Important Note: We have modified the Figures 2 and 3 since original submission for publication in the
Explorer. The reason is that the magnetic and gravity patterns of the Sarasota Arch, off southwest
Florida, do not appear to be offset or interrupted along the projected trace of our "Everglades Fracture Zone". Thus, it may be preferable to consider that SE-ward motion of the Florida Straits block
occurred by sinistral shear jumping northward at the Northeast Basin, which would make that basin a
pull-apart basin for Middle Jurassic time, prior to seafloor spreading in the central Gulf of Mexico.
www.tectonicanalysis.com
105
40
100
95
90
85
80
75
70
40
Simplified from Pindell and Kennan, 2000, in prep.
North America
(present day coordinates)
Max. extent
of Salt Basin
Central
Atlantic
35
South Tamaulipas
Arch now in final
position
Louann
30
Campeche
Yucatán
30
C
Chia
Nazas Arc
(extinct)
Africa
pas
20
35
20
New ridge and
transform
Demarara Rise
South America
Seawater spilled
through here into
the Gulf of Mexico
15
Early Oxfordian 158 Ma
Yucatán-SOAM pole
(Mid. Jur. - E. Cret.)
10
105
100
95
90
85
15
Ocean crust
Marine incursion
ContinentOcean
Boundary
Active
Ridges
80
75
10
70
Figure 3. Late Jurassic (Early Oxfordian) reconstruction of the Gulf
of Mexico and Proto-Caribbean region ("salt fit"). Onset of seafloorspreading stage. Note that Chiapas Massif has been transferred to
Yucatán Block at this time. Also bulk strain direction in Mexico
shifts from ESE-ward to S-ward at this time, with the opening of the
Mexican back-arc basin.
Published in AAPG Explorer, v. 21, October 2000
Important Note: We have modified the Figures 2 to 4 since original submission for publication in the
Explorer. The reason is that the magnetic and gravity patterns of the Sarasota Arch, off southwest
Florida, do not appear to be offset or interrupted along the projected trace of our "Everglades Fracture Zone". Thus, it may be preferable to consider that SE-ward motion of the Florida Straits block
occurred by sinistral shear jumping northward at the Northeast Basin, which would make that basin a
pull-apart basin for Middle Jurassic time, prior to seafloor spreading in the central Gulf of Mexico.
www.tectonicanalysis.com
105
100
95
90
85
80
75
70
Simplified from Pindell and Kennan, 2000, in prep.
North America
(present day coordinates)
40
ry
40
nd
Wiggins Middle
Arch Ground
Arch
Co
nt.
-O
c. B
B1
35
35
A
Mo
jav
Yucatán
on
ora
s
apa
F.
v v
Chi
30
e-S
Espino
Graben
20
Payandé
Trough
15
V
B2
105
100
95
90
30
Guinea
Plateau
Demerara
Rise
Possible extent
of early Salt
Tacatú
Graben
South America
Africa
v v
v
Guyana
Marginal
Fault
?
?
v
vv
vv
Early Jurassic
Transforms
Inactive
Ridges
Poles of
Rotation
Basic
Volcanism
85
20
80
Fracture
Zones
Past position 15
of present
coast
75
70
Figure 4. Early Jurassic reconstruction of the Gulf of Mexico and
Proto-Caribbean region. Onset of "syn-rift" stage.
Published in AAPG Explorer, v. 21, October 2000
Important Note: We have modified the Figures 2 to 4 since original submission for publication in the
Explorer. The reason is that the magnetic and gravity patterns of the Sarasota Arch, off southwest
Florida, do not appear to be offset or interrupted along the projected trace of our "Everglades Fracture Zone". Thus, it may be preferable to consider that SE-ward motion of the Florida Straits block
occurred by sinistral shear jumping northward at the Northeast Basin, which would make that basin a
pull-apart basin for Middle Jurassic time, prior to seafloor spreading in the central Gulf of Mexico.