Recent Faulting and Active Shortening of the Middle Atlas Mountains, Morocco, Within the Diffuse African-Eurasian Plate Boundary

Home , Alpine orogeny, Atlas Mountains, Gibraltar Arc, Middle Atlas

RECENT FAULTING AND ACTIVE SHORTENING OF THE MIDDLE ATLAS MOUNTAINS, MOROCCO, WITHIN THE DIFFUSE AFRICAN-EURASIAN PLATE BOUNDARY

______

A Thesis presented to the Faculty of the Graduate School

University of Missouri-Columbia

______

In Partial Fulfillment

Of the Requirements for the Degree

Master of Science

______

Michael Rigby

Dr. Francisco Gomez, Thesis Supervisor

MAY 2008

The undersigned, appointed by the Dean of the Graduate School, have examined the thesis entitled:

RECENT FAULTING AND ACTIVE SHORTENING OF THE MIDDLE ATLAS

MOUNTAINS, MOROCCO, WITHIN THE DIFFUSE AFRICAN-EURASIAN

PLATE BOUNDARY

Presented by Michael Rigby

A candidate for the degree of Master of Science

And hereby certify that in their opinion it is worthy of acceptance.

Dr. Francisco Gomez

Dr. Robert Bauer

Dr. Brent Rosenblad

ACKNOWLEDGEMENTS

I would like to thank my thesis advisor, Dr. Francisco Gomez, for all of his help throughout the research process. I would also like to thank my thesis committee members, Dr. Robert Bauer and Dr. Brent Rosenblad, for their assistance during the writing process. Also, I want to thank Yousuef Hahou, Zakir Abdelali, Nacer Jabour, and Abdelali Fadil for their assistance during the completion of field work. I would also like to thank the National Science Foundation of the funding of this project (EAR-0408661) and Dr. Rick Allmendinger for the use of his structural programs. Lastly, I would like to thank my family and friends for their continuous support.

ii TABLE OF CONTENTS

ACKNOWLEDGEMENTS...... ii LIST OF FIGURES ...... v LIST OF TABLES ...... vii ABSTRACT ...... viii CHAPTER 1 INTRODUCTION AND TECTONIC BACKGROUND...... 1 Introduction ...... 1 Objectives and Justification...... 4 Tectonic Setting ...... 5 Tectonic Evolution of the Atlas Mountains ...... 6 Triassic through Early Tertiary...... 6 Cenozoic Tectonics ...... 10 CHAPTER 2 - MIDDLE ATLAS STRUCTURE ...... 19 Introduction ...... 19 Geologic Mapping of the Folded Middle Atlas...... 22 Stratigraphy...... 23 Triassic ...... 23 Lower Liassic ...... 23 Middle Liassic ...... 24 Cretaceous ...... 24 Neogene ...... 24 Quaternary...... 24 Remote Sensing Analysis and Interpretation...... 24 Data Types ...... 24 Analyses ...... 27 Stereo Photogrammetry ...... 27 Multi-spectral image processing and interpretation ...... 27 Geologic Map of the Folded Middle Atlas...... 33 Structural Analyses ...... 35 Balanced Cross-Section of the Folded Middle Atlas ...... 40 Discussion...... 42 CHAPTER 3 ACTIVE TECTONISM IN THE FOLDED MIDDLE ATLAS MOUNTAINS...... 45 Longitudinal Stream and Stream Terrace Profiles ...... 46 Quaternary Fault Scarps...... 49 Microtopographic Mapping ...... 49 Radiocarbon Age Dating ...... 57 Fault Scarp Degradation Modeling ...... 58 Single Event Fault Scarp Degradation Model ...... 59 Composite Fault Scarp Degradation Model ...... 66

iii Discussion...... 70 Conclusion ...... 72 REFERENCES CITED ...... 74

iv LIST OF FIGURES

Figure Page

1.1. Simplified tectonic map of the western Mediterranean...... 2

1.2. Topography map of the western Mediterranean...... 3

1.3. Tectonic evolution of the western Mediterranean from 200 - 30 Ma...... 7

1.4. Paleotectonic reconstructions of the western Mediteranean...... 12

1.5. GPS velocity vector map...... 15

1.6. Seismicity Map...... 16

2.1. Geologic map of the Middle Atlas Mountains...... 20

2.2. ASTER band combination image with bands 7-3-1...... 29

2.3. Shortwave infrared spectral reflectance curves...... 30

2.4. ASTER image composed of principal components 4, 5, and 6...... 31

2.5. Geologic map of the Middle Atlas study area...... 34

2.6. Bedding Measurements...... 36

2.7. Exposure of the Ait Oufella Fault in a road cut...... 38

2.8. Fault kinematic data ...... 39

2.9. Balanced cross section...... 41

3.1 Longitudinal Stream and terrace profiles ...... 48

3.2. Field photograph of the composite scarp...... 51

3.3. Field photograph of the singe event scarp...... 52

3.4. 1 meter digital elevation model of the composite scarp...... 54

3.5. 0.5 meter digital elevation model of the single event fault scarp...... 56

3.6. Single event fault scarp profile 1...... 62

v 3.7. Single event fault scarp profile 2...... 63

3.8. Single event fault scarp profile 3...... 64

3.9. Single event fault scarp profile 4...... 65

3.10. Root mean square analysis for D variable...... 68

3.11. Composite fault scarp degradation model ...... 69

vi LIST OF TABLES

Table 2.1 Performance parameters for the ASTER radiometer...... 26

Table 3.1 Radiocarbon age data ...... 57

vii RECENT FAULTING AND ACTIVE SHORTENING OF THE MIDDLE ATLAS MOUNTAINS, MOROCCO, WITHIN THE DIFFUSE AFRICAN-EURASIAN PLATE BOUNDARY

Michael Rigby

Dr. Francisco Gomez Thesis Supervisor

______

Abstract

The intracontinental Atlas Mountains of northwest Africa are prominent tectonic elements in the broader, diffuse plate boundary zone between the converging African and Eurasian plates in the western Mediterranean.

Cenozoic geology, as well as present-day seismicity, document co-eval tectonism across a zone more than 500 kilometers in width. Hence, quantifying the rates of deformation in the Atlas Mountains is necessary in understanding the distribution of crustal deformation across the entire plate boundary zone. This study focuses on constraining the recent shortening rate across the NE-SW striking Middle Atlas Mountains, which are obliquely oriented to the NNW-SSE direction of Late Cenozoic plate convergence.

Both strike-slip and contractional structures are exhibited with apparent slip partitioning: folding and thrusting is concentrated in the Folded Middle Atlas, whereas strike-slip dominates in the Tabular Middle Atlas. In the southwestern part of the Folded Middle Atlas, fault scarps of Quaternary alluvium, including a 4 meter composite scarp and a 1 meter single event scarp, attest to recent faulting along the mountain front. Detailed topographic

viii mapping of the scarps provides a basis for geomorphic analysis and

degradation modeling. Furthermore, the reconstruction of longitudinal stream

terrace profiles helps constrain a long term deformation history. Radiocarbon

dates provide age constraints on the faulted surfaces. To place these active

tectonic observations in a larger context, the fault and fold geometry has been

assessed by completing a 12.5 km structural transect across the frontal

thrust, providing basis for the construction of a balanced cross-section. By

combining the structural geometry with the uplift rate, a minimum estimate of the rate of horizontal shortening in the Middle Atlas is evaluated. Degradation modeling results suggest the Middle Atlas Mountains are shortening at a rate

of 0.03 – 0.05 mm/yr, which is lower than the geodetic estimates. This implies

that shortening at a regional scale is being accommodated by an alternative

mechanism such as strike-slip motion.

ix Chapter 1 Introduction and Tectonic Background

Introduction

The Atlas Mountains of northwest Africa are prominent tectonic

elements in the broader, diffuse plate boundary zone between the converging

African and Eurasian plates in the western Mediterranean. Cenozoic geology, as well as present-day seismicity, document co-eval tectonism across a zone more than 500 kilometers in width extending from southern

Iberia to the edge of the Saharan Platform (Figure 1.1). Hence, quantifying the rates of deformation in the Atlas Mountains is necessary in understanding the distribution of crustal deformation across the entire plate boundary zone.

In that spirit, this study aims to provide the first published constraints on the recent shortening rates across the NE-SW striking Middle Atlas Mountains.

In tectonic terms, the Atlas Mountains are considered a prominent example of an intracontinental mountain belt (e.g., Brede et al., 1992). In contrast to the adjacent Rif and Tel Mountains of North Africa, which comprise the corresponding interplate mountain ranges, the Atlas Mountains are located at a considerable distance into the foreland from their corresponding collision zone (Figure 1.1).

The Atlas Mountains span approximately 2000 kilometers from the

Atlantic coast of Morocco to the Mediterranean coast of Tunisia, and they attain elevations greater than 4 kilometers (in comparison with maximum altitudes of about 2 kilometers in the Rif and Tel Mountains) (Figure 1.2).

Figure 1.1. Simplified tectonic map of the western Mediterranean. Regions surrounding the study area include TMA – Tabular Middle Atlas, HM – High

Moulouya, MB – Missour Basin, and GB – Guercif Basin. The area of Figure

2.1 is outlined by a black box. Figure adapted from Gomez et al. (2000).

Figure 1.2. Topography map of the western Mediterranean. The intracontinental Atlas Mountains attain higher elevations than their corresponding interplate mountains (i.e. the Rif and the Tel)

3 The main ranges that comprise the Atlas are the High Atlas and Middle Atlas in Morocco and the Saharan Atlas in Algeria (Figure 1.1). The NE-SW striking Middle Atlas Mountains are oriented oblique to the general ENE-

WSW strike of the High and Saharan Atlas.

The region adjacent to the plate boundary, including the Alboran Basin and the Rif and Betic mountain ranges, has accommodated the majority of the deformation since the early Miocene (Gomez, 2000). Inland from the plate boundary, the High Atlas, oriented nearly perpendicular to the NNW-

SSE maximum Cenozoic plate convergence, represent a region of compression and have been estimated to have accommodated 17-45% of the overall plate convergence since the early Miocene (Gomez, 2000).

Objectives and Justification

Throughout the intracontinental Atlas Mountains, present day convergence rates and structural geometries are not well constrained. The focus region of this study, the Middle Atlas Mountains, is one key element of foreland deformation occurring within the African- Eurasian plate boundary.

Evidence for recent faulting along the front of the Middle Atlas foreland fold and thrust belt provide incentive for a neotectoinc investigation and an opportunity to incorporate the active tectonic observations with a structural study. In particular, what are the structural styles in the Middle Atlas

Mountains and what is the total Cenozoic shortening? Are there any along

4 strike variations in structural style or shortening? Also, what is the shortening

rate across the front of the Middle Atlas and how does it compare with recent

geodetic estimates? This study intends to evaluate these questions and their

implications on regional tectonics. Furthermore, these objectives not only

have tectonic implications, but also implications for the regional earthquake

hazard. In order to address these objectives, the structure and recent faulting

along the front of the Middle Atlas is evaluated.

Tectonic Setting

Adjacent to the Atlas ranges are two stable crustal blocks that have

experienced little deformation during the Cenozoic: the Moroccan Meseta

located northwest of the High Atlas and the High Plateau located to the

northeast (Figure 1.1) (e.g. Gomez, 2000). To the south of the Moroccan

Atlas are the Anti-Atlas, which are an exhumed Hercynian (late Paleozoic)

mountain chain (e.g. Africa’s Appalachian equivalent) and the stable Saharan

Plateau. Lying to the east and southeast of the Middle Atlas are the Guercif

Basin and Missour Basin. Along the Mediterranean Coast, north of the Atlas ranges, are the western most Alpine orogenic belts: the Rif Mountains and the

Tell Mountains along northern Africa and the Betic Mountains of southern

Iberia. The Alpine mountain ranges in this area are divided into internal allochthonous and external autochthonous zones. The Rif and Betic ranges form an arcuate mountain belt surrounding the Alboran Basin (Figure 1.1).

5 Inherited geological structures likely influence the styles of deformation

in the Atlas Ranges. The present-day Middle Atlas Mountains reflect multiple

phases of deformation including components of extension, compression, and

strike-slip linked to the opening of the Atlantic Ocean and Tethys Sea in the

early Mesozoic and the reactivation of the failed rift basins during Cenozoic

convergence (e. g. Brede et al., 1992; Gomez, 1996). Hence, to place the

Atlas into a broader context, a summary of the Mesozoic and Cenozoic

tectonic evolution of the region follows.

Tectonic Evolution of the Atlas Mountains

Triassic through Early Tertiary

In the late Paleozoic, the supercontinent, Pangea, began breaking up

and forming rift basins as the plates separated. During this time, sedimentary

basins were formed within rift grabens throughout northern Africa (Brede et al., 1992). This represents the first marine incursion of the Tethyan Ocean

over the margin of NW Africa, which resulted in the deposition of a carbonate

platform (Ait Brahim, 2002). The basins formed primarily along two axes

oriented at 30º - 45º or NE-SW and 70º or ENE-WSW, which are parallel to

the Atlantic and Mediterranean trends and correspond to the strike of the

future mountain ranges of the High Atlas and Middle Atlas (Figure 1.3) (Kanes

1973; Brede et al., 1992). Oriented similar to the basins, a system of

synsedimentary faults controlled the development of the basins. Both normal

Figure 1.3. Tectonic evolution of the western Mediterranean from 200 Ma to

30 Ma. Figure adapted from Brede et al., (1992).

7 and oblique slip is observed on the faults in these basins. The existence of both normal and strike-slip components of motion on both fault sets throughout the Triassic suggests a stress field with a least principal extensional stress oriented at approximately NNW-SSE to NW-SE in response to the rifting in the central Atlantic and the opening of the Neo-

Tethys Sea to the east (Brede et al., 1992; Ait Brahim, 2002).

Through the Triassic and into the early Liassic (Lower Jurassic), rifting continued in the basins associated with the High and Middle Atlas ranges.

The stable Moroccan Mesesta separated the Atlantic rifting system from the

Atlas rifting. During the early Jurassic or Liassic, the Atlantic rift system became dominate and the Atlas rift system was aborted. Once aborted, the intracontinental basins in NW Africa began to develop. As the central Atlantic began active oceanic spreading, it initiated a large transform fault zone, the

Newfoundland-Gibraltar fault zone (NGF), due to differential spreading rates of Africa and Iberia (Figure 1.3A). The northern Atlantic remained in a stage of continental rifting causing Iberia to spread at a slower rate than Africa.

Uplift due to rifting diminished in northwestern Africa and the tensional forces from spreading caused regional post-rift thermal subsidence. At this time,

Africa began to move eastward in relation to Iberia due to higher drift velocities along the sinistral NGF.

Throughout the rest of the Jurassic and into the early Cretaceous,

Africa continued to move eastward in relation to Iberia. The African continent remained in relative tectonic quiescence; however, as the continent moved

8 further east, the tensional stresses rotated slightly due to sinistral wrenching

along the NGF (Brede et al., 1992). The tensional stresses within the

continent were oriented at about 125º, which initiated normal faulting along

the 30 º - 45 º striking normal faults and transtensional movement along the

70º striking faults (Figure 1.3) (Brede, 1992).

In the Cretaceous, seafloor spreading began in the north Atlantic, and

Iberia began to slide eastward in relation to Africa causing dextral wrenching on the NGF (Figure 1.3B) (Brede et al., 1992). In addition, a large fault splay within the Newfoundland-Gibraltar fault zone, southwest of Iberia, propagated to form the ALKAPECA microplate (ALKAPECA = Alboran-Kabylian-

Peloritanian-Calabrian, Boulin et al., 1986). This combined movement induced dextral slip along the 70º faults and activated the 30º – 45º faults as oblique reverse faults (Figure 1.3) (Brede et al., 1992). Paleostress analyses estimate a maximum compressive stress oriented E-W to ESE-WNW, which is congruent with the induced movement (Ait Brahim, 2002). This represents the first component of compression within the Atlas; however, this is not nearly as significant compared to the later compression during the Cenozoic.

From the Triassic through the early Tertiary, the differential spreading rates of the Iberian and African plates created zones of weakness within the

African continent that would later be reactivated. The break up of Pangea due to the spreading of the Atlantic Ocean and the opening of the Tethys Sea created transtensional forces throughout northwestern Africa. The primarily

9 extensional deformation through the Mesozoic differs from the compressive

nature of Cenozoic deformation.

Cenozoic Tectonics

During the late Paleocene and early Eocene, the plates began to move eastward at similar velocities (Ziegler, 1988; Brede et al., 1992). Later in the

Eocene, overthrusting began on the borders of the Alkapeca microplate and the emplacement of the Alboren block began in response to the eastward movement of Iberia and Africa (Dewey et al., 1989; Brede et al., 1992).

These tectonic processes caused the first major uplift in the High Atlas and

Middle Atlas that continued on through the early Oligocene. The movement along these fault systems at this time was marked by the uplift and denudation of footwall rocks and the deposition of syntectonic conglomerates.

In the Oligocene, convergence between the continents, linked to the

Alpine Orogeny, continued. The internal zones of the Rif, Betic, and Tell

Mountains were part of an uplifted region corresponding to an area of thickened crust, at least as thick as 40 kilometers based upon gravity anomalies (Platt, 1998). Concurrently, tectonic forces were also transmitted to inland Africa and uplift continued in the Atlas. However, the uplift occurring

in the Atlas during Eocene to Oligocene time was not the major orogenic

event. According to sedimentological studies of the Atlas region, major uplift

and denudation did not begin until the middle Miocene at the earliest (Gorler,

10 1988). Strata within the Missour Basin records uplift within the eastern High

Atlas and Middle Atlas as occurring from the middle Miocene through the

Quaternary (Gomez et al., 2000).

During the middle Miocene, the African and Iberian Plates not only jointly moved eastward but they also converged toward each other (Figure

1.4) (Dewey et al., 1989). However, in contrast to overall plate movement, the thickened crust of the Alboran region began to extend, which dispersed the internal zones of the Rif, Betic, and Tell Mountains. This movement initiated the nappe thrusting of the internal zones of the Rif-Betic Mountains onto the external zones (Brede et al., 1992). At the same time, uplift increased greatly within the High Atlas and Middle Atlas. The faults oriented at 70º, corresponding to the High Atlas, experienced a large amount of up- thrusting; while the faults oriented at 30º - 45º exhibited sinistral transpression.

Since the Miocene, NNW-SSE African-Iberian plate convergence causing thrusting in the High Atlas and transpressive forces within the Middle

Atlas. However, contrary to the Atlas, the thrusting occurring within the Rif and Betic Mountains has been oriented WSW to WNW and concurrent extension has been occurring within the Alboran Basin. This suggests forces acting upon the region other than plate boundary forces. These geodynamic forces are essential to the interpretation of the deformation occurring in this diffuse plate boundary.

Figure 1.4. Paleotectonic reconstructions of the western Mediteranean region. White arrows denote the directions of local tectonic transport. White and black teeth denote inactive and active thrust faults, respectively. The present day coastline is outlined in all time stages. Figure adapted from

Gomez et al. (2000).

12 Three major hypotheses have been suggested to explain the

configuration of the Rif-Alboran-Betic region: subduction rollback leading to

back-arc spreading (Lonergan, 1997), delamination of an overthickened

lithosphere (Seber, 1996, Platt 1998, Fadil, 2006), and convective removal of

mantle lithosphere (Platt, 1998). According to the subduction rollback hypothesis, the oceanic lithosphere must be subducting at such a rate and angle to induce back-arc spreading. In contrast, the delamination and roll

back of overthickened lithosphere induces extension through replacement of

dense lithospheric material with more buoyant asthenosphere. Lastly, the

active removal of mantle lithosphere below the Alboran region would create

high gravitational potential and induce spreading. All three hypotheses do

explain the thinning of the Alboran crust and the existence of extension;

however, new Global Positioning System (GPS) data has provided evidence

to support the roll back of a delaminated subcontinental lithospheric slab

(Fadil et al., 2006).

Based upon GPS data, the present day deformation occurring in the

region is asymmetric and appears to be more indicative of a component of

southward-directed delamination of lithosphere and associated N-S back arc

opening (Fadil et al., 2006). Since the crust of the Rif-Alboran-Betic region is

continental in nature (Platt, 1989), the present day slab is probably the mantle

part of the continental lithosphere, which has become detached from the crust

possibly due to the pull of an old slab (Faccenna et al., 2004, Fadil et al.,

2006). This is further supported by three-dimensional analysis of the upper

13 mantle where very low Pn velocities exist in the uppermost mantle below the

Alboran block and are underlain by a more dense higher velocity zone

(Seber, 1996; Calvert, 2000). The low velocity zone suggests the presence

of a more viscous material underlain by the detached higher velocity

lithosperic mantle. These geodynamic processes incorporated with the plate

boundary forces are responsible for the style of deformation occurring in

northern Africa and the active tectonism in Atlas ranges.

The active setting of the Atlas Mountains can be partially evaluated

through GPS measurements. The GPS measurements give an estimate of

an approximately 5 mm/yr convergence rate between the African and

Eurasian plates in the NW-SE direction (Figure 1.5) (Fadil et al., 2006). In

addition as much as 1 mm/yr convergence is occurring within the Middle

Atlas. However, uncertainties at 95% confidence are generally larger than

the velocities. Hence, GPS can, at best, place a maximum constraint on the

convergence across the Middle Atlas. Therefore, geologic estimates of shortening across the Middle Atlas, which encompass longer time periods, may be more appropriate.

In addition to GPS, recent seismicity throughout the Atlas Mountains

attests to the active nature of the mountain ranges and regional seismic

hazard (Figure 1.6). The most studied event in the Moroccan Atlas occurred

in the western High Atlas in the city of Agadir in 1960. The magnitude of the

event is estimated at mb = 5.9 with an intensity of Io = X for the majority of the

city (Medina et al., 1991; Alami et al., 2004). This event was responsible for a

Figure 1.5. GPS velocity vector map. Instantaneous maximum rate of horizontal shortening of 1 mm/yr for the Middle Atlas Mountains (Reilinger, unpublished).

Figure 1.6. Seismic map indicating earthquake localities with red circles and determined fault plane solutions.

16 large amount of damage and the death of about 12,000 people. Although not as destructive, recent events have also been recorded within the Middle

Atlas. In particular, an event estimated at mb = 4.0 occurred in 1979 along

the frontal thrust fault in the study area (Medina et al., 1991). In addition, nearly 4000 events were recorded throughout northwest Africa and Iberia from 1987 to 1994 including numerous smaller events throughout the High

and Middle Atlas (Alami et al., 2004).

In addition to seismicity, Quaternary volcanism is one of the more

prominent neotectonic features in the area. The timing of volcanism

coincides with the beginning of the present-day regional stress field and the

volcanic centers are approximately aligned NNW-SSE parallel to the

maximum convergence between North Africa and Iberia (Harmond and

Cantagrel, 1984; Ait Brahim, 1984; Morel, 1989; Gomez, 1996). Due to the

alignment of the volcanic centers, it has been suggested that they are related

to crustal scale extensional joints (Harmond and Moukadiri, 1986).

To address the active role of the Middle Atlas Mountains in the diffuse

plate boundary, a study location with identified fault scarps was chosen for an

integrated neotectonic and structural investigation. To begin, a structural

analysis along the front of the mountain range including the completion of a

geologic map and an approximately 12 km balanced cross-section provides

an estimation of shortening and geometry. Next, along the frontal thrust of

the range, fault scarps of Quaternary alluvium are evaluated through detailed

topographic mapping, which provides basis for a geomorphic analysis and

17 degradation modeling. Furthermore, the reconstruction of longitudinal stream terrace profiles helps constrain a long term deformation history and radiocarbon dates provide age constraints on the faulted surfaces. In order to place the active tectonic observations into a larger context, the structural geometry and the uplift rate are combined to give a minimum estimate of the rate of horizontal shortening in the Middle Atlas.

18 Chapter 2 - Middle Atlas Structure

Introduction

In order to quantify the shortening across the southwest portion of the

Middle Atlas, it is necessary to constrain the spatial extent of geologic units

and understand the geometry of the structures in the study region. To

achieve these objectives, a general understanding of the Middle Atlas

structure is necessary along with a detailed geologic map and balanced

cross-section of the study area.

The Middle Atlas Mountains comprise three major geomorphic

provinces: the Tabular Middle Atlas (TMA), the High Moulouya/Missour Basin

area (HMMB), and the Folded Middle Atlas (FMA) (Figure 1.1 and Figure 2.1).

The TMA is the northeastern portion of the Moroccan Meseta consisting of

relatively flat lying Mesozoic strata and the majority of the Quaternary lava

flows. This region is dominated by sinistral strike-slip faulting including the

Tizi-n-Tretten Fault (Figure 2.1) (Gomez et al., 1996). The High

Moulouya/Missour Basin region is located to the southeast of the FMA at the

junction of the High and Middle Atlas and also contains flat lying strata. The

FMA, located between the TMA and the High Moulouya/Missour Basin, is the

focus region of the study and accounts for the large majority of the

compressive deformation in the Middle Atlas. To the northeast, the FMA get as wide as about 55 kilometers corresponding to a major Jurassic rift basin.

Moving toward the southwest, the Folded Middle Atlas is only 20 to 25

kilometers wide.

Figure 2.1. Geologic map of the Middle Atlas Mountains. Labeled structures include AOF – Ait Oufella Fault, TNTF – Tizi-n-Tretten Fault, and NMAF –

North Middle Atlas Fault. Study area is outlined by a black box. Figure adapted from Gomez et al. (1998).

20 These three geomorphic provinces contain variable deformation styles.

The strain in the Middle Atlas Mountains is partitioned with strike-slip motion

concentrated to the northwest in the Tabular Middle Atlas and the majority of

the contraction found in the Folded Middle Atlas (Gomez et al., 1998). In

contrast, the HMMB constitutes the relatively undeformed footwall of the

Middle Atlas frontal thrust.

Bounding the Folded Middle Atlas to the north and separating the

region from the Tabular Middle Atlas is the North Middle Atlas Fault (NMAF)

(Figure 2.1). This fault exhibits primarily sinistral strike-slip motion with a

discontinuous braiding fault trace. Evidence for recent movement along the

fault is provided by faulted Quaternary lava flows and the faulting and folding

of Quaternary deposits (Fedan, 1989; Martin, 1981; Gomez et al. 1996). To

the south of the Folded Middle Atlas are the South Middle Atlas Fault and the

Ait Oufella Fault, which is the frontal thrust of the Cenozoic fold belt in this

region (Figure 2.1). The folds in the region trend primarily NE-SW, similar to

the fault traces, with a secondary E-W trend (Gomez et al., 1996).

Early studies on the development of the Middle Atlas involved wrench faulting and flower structure (Fedan, 1989). More recent studies have suggested thrusting in the Folded Middle Atlas linked to a decollement level

within the Triassic strata (Morel et al., 1993; Gomez et al., 1998). Based on

the structural work located northeast of the study area completed by Gomez et al. (1998), there has been a total of 4.7 kilometers of total shortening across Folded Middle Atlas. The majority of this shortening is accommodated

21 in the southeastern portion of the Middle Atlas toward the Ait Oufella Fault as

over-thrusting.

The study region, which is located in the narrow southwestern portion

of the Folded Middle Atlas, progresses from the northwest toward the

southeast across a series of anticlines and synclines leading up to the active

frontal thrust, the Ait Oufella Fault (Figure 2.1). Geologic mapping, fault

kinematic analysis, and the construction of a balanced cross section has been

completed in order to quantify the shortening across the study region.

Geologic Mapping of the Folded Middle Atlas

Geologic mapping of a transect was completed in order to understand

the structural styles in the southwestern Folded Middle Atlas and to provide field basis for the completion of the balanced cross section. Two primary

methods were used in completing the map: 1) collection of structural

measurements and boundary contacts along a thin transect in the field, and

2) the processing and interpretation of remotely sensed imagery to

extrapolate the field data. Mapping of these data was completed digitally

using ArcGIS Deskstop.

Standard geologic field methods were used to determine orientations

of faults, bedding, and the location of fold hingelines. The Itzer, Morocco

1:50,000 scale topographic map was utilized for determination of

22 measurement localities in the field and as a reference during the completion

of the geologic map.

Stratigraphy

Recognizing stratigraphy and relative ages of the units in the field was

based on previous work in the region (Martin, 1981; Gomez et al., 1996;

Gomez et al., 1998; Itzer Map). The units are divided by age into Triassic,

Lower and Middle Liassic (both within the lower Jurassic), Cretaceous,

Neogene, and Quaternary deposits. Descriptions of these units as seen in the study region follow.

Triassic - The unit includes red argillites, weathered basalts, and

volcaniclastic rocks. The red argillites were the most prevalent rock type of

the Triassic in the study area and served as a distinct marker.

Lower Liassic (Lower Jurassic) - The unit includes massive carbonates,

dolomitic carbonates, and multicolored marls (calcareous mudstone). The

cliff forming massive carbonates are very distinct and are prevalent

throughout the study region.

23 Middle Liassic (Lower Jurassic) – The unit includes sandy carbonates and coral reef deposits. The reef deposits served as the distinctive rock type in this unit.

Cretaceous – This unit lies unconformably on the Jurassic units. It includes microconglomerate, marls and sandy fossiliferous carbonates. The sandy carbonates were the most prevalent rock type of this unit and were typically yellow in color.

Neogene – This unit includes conglomerates and sandy white lacustrine limestones. Outcropping of this unit in the study region is almost completely limited to the footwall.

Quaternary – These deposits include unconsolidated alluvium and colluvium.

Typically, these are found in stream channels, along side the streams as terraces, or alluvial fan deposits.

Remote Sensing Analysis and Interpretation

Data Types Two types of remote sensing data were utilized in completing photo geologic mapping of the study region: aerial photographs and Advanced

Spaceborne Thermal Emission and Reflection Radiometer (ASTER) imagery.

24 These data were used for stereoscopic viewing and lithologic mapping of the study region.

Aerial photographs at 1:60,000 scale were utilized prior, during, and after field work was completed to assist in terrain interpretation and mapping.

The photographs were obtained in a 1973 flight mission and coverage was limited to the front portion of the study region.

The ASTER sensor is one component of the Earth Observation

System (EOS) TERRA satellite that was launched in 1999 jointly by the

United States and Japan. The ASTER sensor records solar energy in 14 spectral bands through 4 different telescopes. The sensor measures three bands in the visible near infrared (VNIR) with an additional back-looking telescope for the generation of digital elevation models (DEM) at a 15 meter spatial resolution. It measures six bands in the shortwave infrared (SWIR) at

30 meter resolution and five bands in the thermal infrared at 90 meter resolution (Table 2.1). The VNIR and SWIR bands were used for this study.

The thermal bands were not used. PCI Geomatica OrthoEngine was used for primary processing and generation of the DEM, while ER Mapper was utilized for further processing. Table 2.1 outlines the performance parameters for the

ASTER radiometer taken from Fujisada (1995).

the ASTER radiometer. Table from Table 2.1 Performance parameters for Fujisada (1995).

26 Analyses

Stereo Photogrammetry

Stereo viewing of aerial photographs was used to aid in the production of the geologic map by extending rock unit contacts and interpreting

Quaternary deposits from above. In addition, it provided three-dimensional perspective of the study area and an additional view of the landscape to incorporate with the satellite imagery.

The ASTER imagery was also processed for stereoscopic viewing.

The stereographic bands (3N and 3B) were used to produce a digital elevation model (DEM). Generation of the DEM was completed in order to produce a higher resolution topographic image than the existing Shuttle

Radar Topography Mission (SRTM) dataset, which has a resolution of 90 meters. The DEM produced from the ASTER imagery has a 15 meter spatial resolution revealing a much more detailed topography.

Multi-spectral image processing and interpretation

Lithologic mapping of the study region was completed by constructing false-color images from the multi-spectral ASTER bands (e.g. bands 1-9 for this study). Multiple processing approaches were utilized, including the combination of different bands, creating band ratios, and principal components analysis, in order to optimize the usefulness of the imagery.

After trying multiple combinations, two variations of the ASTER image were

27 created for the identification of the lithologic units. The first is a combination of bands 7-3-1 as a red-green-blue (RGB) image (Figure 2.2). The second is a combination of principal components 4, 5, and 6 (PC4-5-6) as an RGB image (Figure 2.4).

The 7-3-1 image combines shortwave infared, near infared, and visible bands (Figure 2.2). Similar images have been used by Pena (2004) for oil and gas exploration. This combination has the advantage of preserving morphological features as well as displaying lithological units in different colors (Pena, 2004). The SWIR bands are useful here since many sedimentary rocks have distinctive spectral characteristics in this portion of the electromagnetic spectrum (Evans, 1988). For instance, limestone has higher reflectance than shale throughout the SWIR (Figure 2.3) (Pena, 2004).

This image was very useful for extending unit contacts wherever bare rock was visible (Figure 2.2). This image, however, does not clearly decipher units when large amounts of vegetation are present. This is due to the high reflectance of vegetation in the near infrared. Therefore, an additional image that is not affected by vegetation is needed for the areas within the study region covered in vegetation.

The PC4-5-6 image displays principal components 4, 5, and 6 as an

RGB image (Figure 2.4). The principal components analysis combines all nine bands into a nine dimensional analysis. The analysis begins by shifting the origin of a nine dimensional plot to the point representing the mean of all

Figure 2.2. ASTER band combination image with bands 7-3-1 (RGB). Study area outlined on Figure 2.1. Yellow lines are lithologic contacts.

Figure 2.3. Shortwave infrared spectral reflectance curves for limestone and shale. Location of the ASTER bands are shown in gray. Figure adapted from

Pena (2004).

Figure 2.4. ASTER image composed of principal components 4, 5, and 6

(RGB). Study area outlined on Figure 2.1. Yellow lines are lithologic contacts.

31 the bands. The next step creates axes along the line of greatest spread to create PC1 (the principal axis of an ellipse in 2 dimensional analyses). The analysis continues by creating principal components for all nine dimensions

(Drury, 2001). The principal components 4, 5, and 6 were used here to eliminate the axes of highest correlation (e.g. topography and vegetation).

The effects can be observed in Figure 2.4 as topography can not be distinguished and the regions with large amounts of vegetation (e.g. the stream valleys) are black. This image was particularly useful in displaying

Triassic deposits, which are clearly shown in purple in the NW portion of

Figure 2.4.

The photo geologic mapping of the study region was completed utilizing both stereo photogrammetry and processed multispectral ASTER images. By utilizing imagery, the contacts between the rock units established in the field were able to be extended. In addition, 3 point attitudes provided additional topographic constraints (e.g. Itzer, Morocco topographic map) on the mapped contacts and the extracted DEM, created from bands 3N and 3B, was useful for displaying the topography of the study region.

32 Geologic Map of the Folded Middle Atlas

By combining the field data, stereographic photo interpretation, and the multispectral ASTER imagery, the geologic map of a portion of the Folded

Middle Atlas was created (Figure 2.5). The map region is approximately 13 kilometers by 5 kilometers and is oriented NW-SE parallel to the cross section line and the maximum Cenozoic shortening direction in the Middle Atlas.

Structural trends throughout the mapped region are generally oriented

NE-SW with secondary E-W fold trends. Major structures include the Ait

Oufella frontal thrust, a syncline, an anticline, a NW dipping fault, and a SE dipping back thrust. The major syncline and anticline are open with shallow limbs and extend throughout the narrow portion of the Folded Middle Atlas toward the NE. Within the study area, the hingeline of the syncline rotates from a NE-SW trend to an E-W trend. This arcuate hingeline is likely due to the bend in the AOF just to the SW of the mapped region since its trace nearly parallels the fault trace. The back thrust is also a prominent feature in the region mapped in other studies (e.g. Martin, 1981; Gomez et al., 1998).

However, the NW dipping fault seems to be a smaller splay that terminates into the back thrust just outside the mapped region. A minor anticline and syncline occur near the frontal thrust and terminate within the study region

SW of the cross-section line (Figure 2.5). A small oblique dextral thrust fault is located in the hinge of the anticline. These structures may also be in response to the bend in the frontal thrust. In addition, smaller undulations in

Figure 2.5. Geologic map of the Middle Atlas study area. Scarp localities are shown as stars. Balanced cross section line extends from A-A’.

34 stratigraphy occur throughout the mapped region; however, these are not

significant to the larger structure.

Structural Analyses

Field data, including bedding attitudes and fault kinematic data provide

basis and justification for the balanced cross-section. Two analyses were

performed including stereonet analysis of bedding measurements in the study

region and a fault kinematic study of structural measurements along the Ait

Oufella frontal thrust. FaultKin © and StereoWin © were used to analyze and

plot the data.

Stereonet analysis of bedding orientations throughout the study region

was completed in order to determine general fold axis orientations and the

finite strain axis. The poles to bedding were plotted on an equal area

stereonet. The best fit plane and its pole were determined and also plotted

on the same stereonet (Figure 2.6). The trend of the bedding poles (319º)

corresponds to the finite strain axis and reaffirms the principal NW-SE

transport direction through the southwestern portion of the Folded Middle

Atlas. The orientation of the pole to the best fit plane is an estimation of the fold hingelines in the study region. Data generally fit well for a cylindrical fold.

This is an essential assumption used in the construction of the cross-section that follows since cylindrical folding is required to eliminate mass transport into and out of the plane of section.

bedding. Strike of the cylindrical best fit plane to the finite strain axis. Line denoted by point 3 in red is a Figure 2.6. Equal Area stereonet plot of poles to depicted by the red line is parallel general estimate of hingeline orientation.

36 Fault kinematic data were collected in a road cut along the Ait Oufella frontal thrust (Figure 2.7). The data are plotted as faults and striations with their associated Bingham axes and as a summed fault plane solution (Figure

2.8). The fault plane solution represents the unweighted geometric moment tensor summation (Marrett and Allmendinger, 1990). The northwest dipping plane (211º, 30º) corresponds to the Ait Oufella Fault. This orientation is congruent with earlier fault kinematic analyses (e.g. Gomez et al., 1996;

Gomez et al., 1998) and attests to a NW to SE transport direction. The lack of significant oblique movement along the fault plane supports the assumption of two-dimensional movement within the balanced cross section plane. In addition, the infinitesimal strain axes are co-axial with the finite strain axes

(cylindrical best fit; Figure 2.6) indicating pure shear through the study region.

This is an important observation since the co-axial strain, in addition to cylindrical folding, are assumed in the construction of the balanced cross section that follows.

Figure 2.7. Exposure of the Ait Oufella Fault in a road cut. Triassic Argillite is being thrust over the top of Neogene conglomerate.

Kinematic data were collected plots of fault kinematic data. Figure 2.8. Equal area stereonet from the road cut in Figure 2.7.

39 Balanced Cross-Section of the Folded Middle Atlas

Based upon the structural analyses, the overall structural style across

the Folded Middle Atlas appears to be two dimensional with cynlindrical

folding and co-axial strain. Hence, the construction of a balanced cross section across the southwestern portion of the Folded Middle Atlas can place a minimum constraint on the shortening across the fold belt. The section line extends 13 kilometers from the High Moulouya platform across the AOF to a back thrust and is oriented N45ºW-S45ºE (e.g. parallel to the axis of principal shortening) (Figure 2.6).

At the frontal thrust, Triassic beds are oriented parallel to the fault surface inferring layer parallel shear (Figure 2.9). In addition, deformation in the study region is brittle in nature and fold geometries appear to be related to bends in the fault surface (i.e. fault-bend folding). Within the cross section, folding was generally assumed to be concentrated along “kink axes”, which follow the fault-bend folding style outlined by Suppe (1983) (Figure 2.9).

Stratigraphic thicknesses throughout the cross section were constrained by rock unit contacts within the study region and earlier studies of footwall thicknesses (Martin, 1981; Lorenz, 1976).

The structural interpretations of the cross section have been balanced using FaultFold ©. This program assumes parallel folding through conserving mass by maintaining equal bed lengths and thicknesses throughout deformation. The program was utilized to forward model the deformation by

40 outhwestern portion of the Folded Middle Atlas Figure 2.9. Balanced cross section of the s Mountains.

41 adjusting initial parameters (e.g. fault geometries and total slip) to fit the observed deformation. By restoring the beds in the study region to an undeformed state, an estimate of 3.9 kilometers of total Cenozoic shortening across the Ait Oufella Fault was determined (Figure 2.9).

Discussion

The “thin-skinned” structural style depicted in the cross section (Figure

2.9) is suggested by field observations. First, Triassic beds in the hanging

wall lie parallel with the AOF indicating a detachment surface. Second, in a

thick-skinned model, where the fault extends downward at a steeper angle

into the basement, the amplitude of the fault bends would increase.

According to the basic kinematics of fault bend folding, larger amplitude fault

bends will be compensated by steeper fold limbs (Suppe, 1983). Hence, a

low-angle fault geometry seems more appropriate for the study area.

For the Middle Atlas, the typical “layer cake” geometry associated with

thin skinned thrust belts must be slightly adjusted since this deformation is

superimposed on a zone of earlier extension. Within the hanging wall, lower

Jurassic (Liassic) strata are prevalent; however, they are not present in the

footwall of the AOF (i.e. the High Moulouya). Instead, Triassic strata are unconformably overlain by Cretaceous strata (Martin, 1981). This implies a

paleogeographic boundary between the two regions, such as a Mesozoic normal fault (Figure 2.9). Based on stratigraphic discordance between the

42 hanging wall and footwall of the AOF, this normal fault bound a major

Jurassic rift basin. Though, it should be said that this steep basement fault is

not reactivated during Cenozoic convergence due to geometric constraints

provided by the fold geometries previously described. However, basement

structure may play a role in features such as the lobe-like front of the Middle

Atlas and the shape of the syncline at the front.

Straitigraphic thicknesses throughout the study area are generally uniform; however, the Middle Liassic unit does pinch out toward the southeast

into angular unconformity with the Cretaceous (Figure 2.9). Generally, this

contact is disconformable or in slight angular unconformity, although, more

pronounced angular unconformity has been documented (Ensslin, 1992;

Gomez et al., 1998). The existence of the angular unconformity in the

restored section of the study area suggests an initial slope of approximately

3º for the Triassic and Jurassic strata in the hanging wall. This initial dip

could be a result of tilting coincident with Mesozoic extension in the basin.

A more regional objective for this study was to determine if any along

strike variation in structural style and shortening occur along the Ait Oufella

Fault. In a study by Gomez et al. (1998), located about 25 kilometers

northeast of the study area in the Folded Middle Atlas, similar thin-skinned

deformation was observed with a total of 4.7 ± 0.7 kilometers of shortening.

The determined shortening from this study (i.e. 3.9 kilometers) is less;

however, the cross section from Gomez et al. (1998) covers a larger distance

and, therefore, contains more shortening. Regardless of the shortening

43 discrepancy, the similar structural styles provide collaborative data to support

a thin-skinned deformation model for the southwestern portion of the Middle

Atlas Mountains.

Within the study region, the Triassic, composed of relatively “weak”

rocks (e.g. argillite, evaporates, and weathered basalts), serves as the

detachment layer for the thin-skinned deformation. However, progressing

towards the northwest, the detachment merges with left lateral faults (e.g. the

North Middle Atlas Fault and the Tizi-n-Tretton fault) (Figure 2.1). The

geometry at depth at the junction of these faults is not well constrained and can not be inferred from this study.

Regardless of the structure at depth, it is important to link these structural observations and results to the active faulting occurring in along the front of the Middle Atlas. In the section that follows, the active setting of the

Middle Atlas Mountains is evaluated in order to determine a shortening rate across the Middle Atlas.

44 Chapter 3 Active Tectonism in the Folded Middle Atlas Mountains

In order to understand the recent deformation and quantify the shortening rate across in the southwestern portion of the Folded Middle Atlas,

a neotectonic investigation of the study region is necessary. Evidence for

recent faulting along the Ait Oufella frontal thrust is evident due to faulted

stream terraces, terrace and stream profile inflections, and recent seismicity.

An mb = 4.0 earthquake was recorded in 1979 along the Ait Oufella frontal

thrust (Medina et al., 1991). However, the earthquakes throughout the Middle

Atlas are typically quite small and consequently have poorly constrained

epicenter locations. Therefore, investigating the neotectonic geomorphology

(e.g. faulted terraces and stream and terrace profiles) along the frontal thrust

provides valuable information for interpretation of the active deformation in

the study region. In addition, this investigation provides data to quantify the

preliminary uplift rate and, in conjunction with the cross-section, the

shortening rate across the southwestern Folded Middle Atlas.

The primary objectives of the neotectonic investigation were to 1)

reconstruct longitudinal stream and terrace profiles to help constrain a long term deformation history and 2) survey the fault scarps to provide basis for

geomorphic analysis and degradation modeling. In addition, radiocarbon

dates provide age constraints on the faulted surfaces.

Collection of field data was completed using real time kinematic (RTK)

GPS. This land surveying technique utilizes the carrier frequency in the GPS

45 signal to provide real-time corrections to centimeter accuracy. A base station

receiver broadcasts the phase of the carrier that it measured from the satellite

to a mobile unit, and the mobile unit compares that with the phase that it

received. This allows the relative position of the units to be determined with

centimeter precision; although, the absolute accuracy is a function of the accuracy of the base station location. RTK GPS points were collected on numerous stream terrace surfaces and on the faulted surfaces.

Longitudinal Stream and Stream Terrace Profiles

The longitudinal stream and stream terrace profiles were created in order to constrain a deformation history across the Ait Oufella frontal thrust.

Profiles spanning nearly 2 kilometers across the AOF were produced to

observe tectonic related inflections (Figure 3.1). The present day stream

profile was created using the Itzer, Morocco 1:50,000 scale topographic map

and RTK GPS points, while terrace profiles were produced using RTK GPS

points only. Measured terrace surfaces are represented on Figure 3.1 as

solid lines with interpretive dashed lines projected between the measured

surfaces. Measured surfaces through the profiles are limited due to stream

erosion.

The inflection of the present stream profile and the terrace level 1

profile is evident and attests to the recent activity along the frontal thrust

(Figure 3.1). The Ait Oufella Fault cuts directly though terrace level 1 forming

46 a composite fault scarp that is discussed in much more detail in the section

that follows. The present-day stream profile contains a large inflection nearly

a half kilometer upstream from the fault zone. This is possibly due to the

active erosion occurring in the stream. As uplift continues at the front, the

upstream gradient will increase along with topography, and, therefore, the

erosive power of the stream will increase causing the inflection to propagate

upstream. The remnant terrace level 1 is not actively eroding, and therefore, the inflection at the fault is preserved (i.e. the fault scarp). Another possible cause for this inflection is a change in lithology. Near the inflection is the

Triassic/Jurassic boundary where Triassic weathered basalts and argillites are overlain by massive cliff-forming carbonate. The differential weathering

between these layers could cause an inflection in the stream profile. Another

factor that may affect the shape of the present stream profile is the contour

interval of the topographic map used in its construction. The majority of the

profile is based off of a topographic map with a resolution of 20 meters.

Therefore, the exact location of the inflection could shift slightly toward the

fault zone.

Terrace surfaces 2 through 4 were not continuous across the fault

zone; therefore, inflections in their profiles were not observed. However, by projecting the existing data, an increase in the slope of these surfaces across the fault zone is required. Typical, undeformed stream terrace profiles would

also expect to steepen progressively upstream; therefore, the extent of the

tectonic influence on these surfaces is not well constrained.

and dip of AOF are iles. Solid lines denotes iles. inferred. Location eam and stream terrace prof Figure 3.1. Longitudinal str accurate. measured surfaces, while dashed lines are measured surfaces, while

48 Quaternary Fault Scarps

Along the Ait Oufella Fault, 1 meter and 4 meter fault scarps of

Quaternary alluvium provide basis for geomorphic analysis and degradation

modeling (Figure 3.2 and Figure 3.3). The 1 meter scarp is inferred to be a result of a single faulting event, while the 4 meter scarp, located in terrace level 1, is inferred to be a composite scarp. Fault scarps were surveyed using

RTK GPS. High resolution digital elevation models of the scarp forms provide topographic perspective and justification for the placement of perpendicular profiles for degradation modeling. Radiocarbon data provide age constraints on the faulted surfaces and required age estimates for the modeling. By combining radiocarbon age data with degradation modeling, an uplift rate and, therefore, a shortening rate can be determined.

Microtopographic Mapping

The topography of the fault scarps was investigated using high resolution digital elevation models in order to provide an overhead perspective of the scarp forms and to identify appropriate modeling profiles.

The DEMs were created from thousands of RTK GPS points collected over the entire area of the scarps. ArcInfo Workstation was used to create the

DEM by interpolating between the data points. A 1 meter resolution DEM was created for the composite scarp and the profile was chosen along the most representative transect (Figure 3.4). A 0.5 meter resolution DEM was

49 created for the single event fault scarp and four perpendicular profiles were chosen for degradation modeling (Figure 3.5). ArcScene was utilized to

create three-dimensional perspectives of the fault scarps based on their

DEMs (Figure 3.4 and Figure 3.5).

Figure 3.2. Field photograph of the composite scarp. The truck is on the top surface of the scarp, while the people are on the bottom. The photograph was taken facing west.

Figure 3.3. Field photograph of the singe event scarp. Relative displacement and scale can be assessed from the backpack and camera case. Photograph taken facing northwest.

53 Figure 3.4. 1 meter digital elevation model of the composite scarp. Scarp profile used for degradation modeling is indicated with solid black line.

55 Figure 3.5. 0.5 meter digital elevation model of the single event fault scarp.

Scarp profiles used for degradation modeling indicated with solid black lines.

56 Radiocarbon Age Dating

Radiocarbon dating of the faulted surfaces was completed to provide age constraints for the determination of an uplift rate. Two types of samples

were collected from within the faulted surfaces for age dating: individual

charcoal fragments and bulk sediment. The charcoal fragments were sent to

the University of Arizona, while the bulk sediment sample was sent to the

University of Georgia.

The ages received after analyses were reported as radiocarbon ages

before present (B.P.) and can not be utilized without calibration. In order to

determine a calibrated calendar age, the “raw” ages must be calibrated

against an atmospheric carbon curve since the amount of carbon in the

atmosphere has varied with time. The ages were calibrated to calendar years

using Oxcal calibration software (Bronk-Ramsey, 1998). The calibrated ages

in years B.P. are displayed in Table 3.1 as confidence intervals to one and two standard deviations.

Scarp Sample Type Radiocarbon Age δ13C Cal Age (2 σ) Single Event Alluvial Charcoal 3860 ± 41 BP 22.6 2470 BC - 2200 BC Single Event Alluvial Charcoal 3831 ± 55 BP 21.1 2470 BC - 2130 BC Composite Bulk Colluvium 4480 ± 30 BP 22.8 3340BC - 3020BC

Table 3.1. Radiocarbon age data

57 The age dates for the single event scarp provide a maximum age for

the last faulting event along the AOF. The age date from the composite scarp, however, is too young to represent the age of the faulted terrace level

1. It is more likely that this is the age of colluvial wedge material deposited in the front of the scarp after a seismic event. This age is useful since it does reaffirm recent faulting occurring along the frontal thrust. However, the uplift rate associated with the single event scarp (e.g. 1 meter per 4300 years) is not constrained and only gives an estimate based on the last surface rupturing event in this area. Therefore, another method must be used to estimate the age of the composite scarp in order to determine a more representative and longer term uplift rate for the Ait Oufella fault.

Fault Scarp Degradation Modeling

The fault scarp degradation modeling was completed in order to estimate the age of the composite scarp and determine the uplift rate along the Ait Oufella Fault. Modeling of both scarps was needed in order to obtain these estimates and determine a shortening rate across the AOF.

The degradation modeling completed for this study utilizes the diffusion equation analysis outlined by Hanks et al. (1984, 2000). This analysis inherently contains two physical assumptions. The first is that mass is conserved at a local scale, and the second is that erosional diffusivity is not much different than depositional diffusivity, which allows for the assumption of

58 a spatially constant mass diffusivity, k (Hanks, 2000). An additional assumption is that mass moves downhill at a rate proportional to du/dx and that k is not time dependant. Based on these assumptions, a one- dimensional, homogeneous diffusion equation with constant coefficients can be represented as:

du d 2u − k = 0 (1) dt dx 2

The relative elevation, u, is a function of profile distance, x, and time, t. In other words, the rate of change in elevation is proportional to its curvature

(Hanks, 2000). Therefore, a region along a profile that is concave up or has positive curvature will have u increase with time and vise versa. For purposes of this study, k is represented in meters squared per kiloannum (m2 ka-1), while time is represented in ka.

Single Event Fault Scarp Degradation Model

The degradation modeling of the single event fault scarp was completed in order to determine a representative mass diffusivity constant, k, for the study region. This value is required to determine of the age of the composite scarp and the uplift rate along the AOF. For a better estimate of k, four profiles were chosen perpendicular to the trend of the scarp (Figure 3.5). The points for each profile were extracted from the RTK GPS points along a 1 meter wide swath.

59 To model the single event scarp degradation, an analytical solution to equation 1 can be implemented for the case of a single vertical step in topography of 2a imposed at t=0 upon a surface of pre-existing slope b

(Hanks, 1984). This corresponds to a single vertical dip slip faulting event of a surface with an initial slope of b. According to Hanks (1984), the solution to equation 1 with these initial conditions is:

⎛ x ⎞ u(x,t) = a erf ⎜ ⎟ + bx (2) ⎝ 2 kt ⎠

The profile distance x is referenced from an origin at the scarp mid-height.

This produces an uplift/down drop of ± a at x=0 for time, t, instantaneously greater than zero.

For unconsolidated materials, such as the material that comprises the faulted surfaces, an initial vertical offset is somewhat unrealistic. In reality, the scarp would degrade almost immediately to the angle of repose for that specific material. However, due to the large amount of curvature induced by the initial offset, the scarp will degrade at a much faster rate initially. For the time scales involved in this degradation study, the inaccuracy of the initial condition will be muted.

Utilizing equation 2, the single event scarp profiles were modeled

(Figure 3.6 through Figure 3.9). Model variables (e.g. k, a, and b) were adjusted to best fit the points for each profile. Variable values of a and b are reasonable due to along strike variations in initial slope and the termination of the scarp to the NE of profile 4 (Figure 3.5 and Figure 3.9). An age of 4300

60 B.P. (t = 4.3) based on the radiocarbon dates was used for all four profiles.

Results are shown on Figures 3.6 through 3.9.

The mass diffusivity, k, was determined from the four modeled profiles to range from 0.9 – 1.5 m2/ka. This is a reasonable range of values for a semi-arid region such as Morocco. Similar values were determined by Hanks et al. (1984) for regions within the Basin and Range Province (i.e. 1.1 m2/ka).

along profile 1 rp degradation model Figure 3.6. Single event fault sca (Figure 3.5)

along profile 2 rp degradation model (Figure 3.5) Figure 3.7. Single event fault sca

rp degradation model along profile 3 Figure 3.8. Single event fault sca (Figure 3.5)

along profile 4 rp degradation model (Figure 3.5) Figure 3.9. Single event fault sca

65 Composite Fault Scarp Degradation Model

In order to determine an uplift rate along the AOF, degradation

modeling was completed for the composite fault scarp. Profile points were extracted from the RTK GPS points along a 5 meter swath. Since the solution used for the single event scarp is not applicable for a multiple event

scarp, a separate solution must be derived. Hanks et al. (1984)

parameterizes faults that have experienced multiple episodes of faulting in

terms of a slip rate, 2A. The result is an adjustment to Equation 1 in the form

of:

du d 2u − k = A (3) dt dx 2

The solution to this equation shown below is for an initial scarp offset of ± a at

x >and< 0 cut on a preexisting surface of slope b. For x > 0, A corresponds to

uplift, and for x < 0, A corresponds to downdrop so that u = 0 and x = 0 for all

2 2 ⎡ ⎤ −x ⎛ x ⎞ Ax ⎛ x ⎞ Ax kt 4kt u(x,t) = ()a + At erf ⎜ ⎟ + ⎢erf ⎜ ⎟ − sgn()x ⎥ + e + bx (4) ⎝ 2 kt ⎠ 2k ⎣ ⎝ 2 kt ⎠ ⎦ k π

where

sgn(x) = +1 x > 0

sgn()x = −1 x < 0

and a and A are now positive for all x (Hanks et al., 1984).

66 Further manipulation to this solution can be done since the existing

composite scarp height is known from the RTK GPS points. The updated

solution is of the form:

2 ⎛ x ⎞ Qx 2 ⎡ ⎛ x ⎞ ⎤ Q D −x u(x,t) = ()a + H erf ⎜ ⎟ + ⎢erf ⎜ ⎟ − sgn()x ⎥ + e 4D + bx (5) ⎝ 2 D ⎠ 2 ⎣ ⎝ 2 D ⎠ ⎦ k π

where

H = At D = kt A Q = k

Since one half the slip rate, A, multiplied by the age, t, renders one half the

height of the scarp, H is a known constant. By assigning three new variables

to the equation, the uplift rate and the age of the composite scarp can be solved for simultaneously by holding H constant and statistically determining the best combination of D and Q for the given scarp profile.

A root mean square analysis was performed to calculate a best fit H for the profile data points. This value was then entered into the model and a root mean square analysis was completed to determine the best fit values for D and Q (Figure 3. 10). Based on the statistical analysis, a best fit D value of

200 and Q value of 0.0095 were determined with the initial conditions of H =

1.9 meters and b = 0.020 (Figure 3.11). By entering the range k values determined from the single event scarp analysis, the age of the composite scarp is estimated to range from 130 – 220 ka. In addition, based on these values, the uplift rate along the Ait Oufella Fault is estimated to range from

0.02 – 0.03 m ka-1.

Figure 3.10. Root mean square analysis for D variable in composite fault scarp degration model. Best fit value D = 200.

Figure 3.11. Composite fault scarp degradation model along profile line on

Figure 3.4.

69 By utilizing the dip of the fault (i.e. 30º), uplift rate can be correlated to

a rate of horizontal shortening. The shortening rate across the Ait Oufella

frontal thrust, determined from the uplift rate, is estimated to range from 0.03

– 0.05 m ka-1 or 0.3 – 0.05 mm/yr.

Discussion

GPS measurements of instantaneous convergence rates across the

Middle Atlas Mountains provide a maximum estimate of horizontal shortening

of approximately 1 mm/yr (Fadil et al., 2006). However, the 95% confidence

limits on the GPS velocities are similar to the velocities themselves, so the

GPS measurements do not constrain well the shortening across the Middle

Atlas. The current study determined a much lower estimate of 0.03 – 0.05

mm/yr for late Quaternary shortening across the frontal thrust, which could be a result of how the strain is being accommodated. Even though the Ait

Oufella Fault is the primary frontal thrust of Cenozoic shortening, more recently, strain may be accommodated by other structures in the region (e.g. strike slip faults, blind thrusts, folds). This would account for the discrepancy between the two data values.

In addition to strain accommodation, the modeled degradation age contains many possible sources of error including some of the fundamental assumptions of mass conservation at a local scale. The scarp surface has undergone minor cultivation, and, depending on the extent of this cultivation,

70 this could have an effect on the age estimate. However, more likely, the

diffusivity value, estimated from the single event scarp, may be underestimated. This value has a large effect on the calculated age. By increasing the diffusivity value to 5.0 m2 ka-1 the age of the scarp is reduced to 40 ka with a corresponding increase in the shortening rate to 0.16 mm/yr.

However, regardless of the estimate of k, accommodation of strain by other structures is warranted since increasing k to an unrealistic value of 15 m2 ka-1 only results in a shortening rate of 0.49 mm/yr.

In addition to the shortening rate, the modeled composite scarp age has implications on earthquake hazard. Based on the single event scarp height (i.e. about 1 meter), it can be deduced that the composite scarp is the result of four earthquake events. By combining this information with the age of the scarp, the recurrence interval is estimated to range from 33 – 55 ka.

With a maximum age constraint of approximately 4300 years on the last

faulting event, the earthquake hazard for this region is moderate to low.

However, these results are preliminary and further age constraint on the

composite scarp is needed for a more certain estimate of recurrence interval.

Additionally, from the single event scarp height, an estimate of the

magnitude for the seismic event associated with the formation of the single

event scarp can be determined. The magnitude is estimated using the

empirical relationship between coseismic displacement and magnitude

established by Wells and Coppersmith (1994). Based on the height of the

scarp (i.e. 1 meter) and the dip of the AOF (i.e. 30 º), the displacement

71 associated with this seismic event is 2 meters. Placing this value into the

empirical relationship for the maximum displacement, an estimate of M = 6.7

provides a minimum constraint on the magnitude. Alternatively, the empirical

relationship for the average displacement gives a similar estimate of M = 6.7.

Therefore, there is the potential for a large destructive earthquake along the

Ait Oufella Fault despite the long recurrence interval.

Conclusion

The NE-SW trending Middle Atlas Mountains are an active

intracontinental mountain belt within the diffuse African – Eurasian plate

boundary. The Middle Atlas study area in the southwestern portion of the

Folded Middle Atlas contains fault scarps of Quaternary alluvium that provide basis for the determination of the horizontal shortening rate from degradation modeling. The late Quaternary shortening rate across the Ait Oufella frontal thrust of 0.03 – 0.05 mm/yr determined in this study in much lower than the maximum estimate of 1 mm/yr provided by GPS measurements. Most likely,

this discrepancy is due to the compressive strain within the region being

accommodated by structures other than the AOF.

To place these active tectonic observations in a larger context, the fault

and fold geometry was assessed by completing a 12.5 km balanced cross

section across the frontal thrust. From the cross section, a minimum estimate

of total Cenozoic shortening of 3.9 kilometers was determined. This structural

72 investigation provided basis for an interpretation of thin-skinned deformation occurring along the front of the southwestern Middle Atlas. This correlates well with other, along strike, studies (i.e. Gomez et al., 1998).

This study integrated structural, neotectonic, and geomorphic analyses to investigate Middle Atlas foreland deformation. The knowledge gained from this investigation provides the first geologic estimate of shortening across the

Middle Atlas Mountains of Morocco. In addition, the evaluation of the

structural geometry for the southwestern Middle Atlas front offers new data to

assist in understanding the style of deformation occurring within the diffuse

African-Eurasian plate boundary.

73 References Cited

Ait Brahim, L., and Chotin, P., 1984, Mise en évidence d'un changement de direction de compression dans l'avant-pays rifain (Maroc) au cours du Tertiaire et du Quaternaire: Bulletin de la Societé Geologique du France, v. 7, p. 681-691.

Ait Brahim, R.L., Chotin, P., Hinaj, S., Abdelouafi, A., El Adraoui, A., Nakcha, C., Dhont, D., Charroud, M., Alaoui, F.S., Amrhar, M., Bouaza, A., Tabyaoui, H., and Chaouni, A., 2002, Paleostress evolution in the Moroccan African margin from Triassic to Present: Tectonophysics, v. 357, p. 187-205.

Brede, R., Hauptmann, M., and Herbig, H.-G., 1992, Plate tectonics and the intracratonic mountain ranges in Morocco -- The Mesozoic-Cenozoic development of the Central High Atlas and the Middle Atlas: Geologische Rundschau 81/1, p. 127-141.

Bronk-Ramsey, C., 1998, Probability and dating: Radiocarbon, v. 40, p. 461- 474.

Calvert, A., Sandvol, E., Seber, D., Barazangi, M., Roecker, S., Mourabit, T., Vidal, F., Alguacil, G., and Jabour, N., 2000, Geodynamic evolution of the lithosphere and upper mantle beneath the Alboran region of the western Mediterranean: Constraints from travel time tomography: Journal of Geophysical Research, v. 105, p. 10871-10898.

Dewey, J.F., Helman, M.L., Turco, E., Hutton, D.H.W., and Knott, S.D., 1989, Kinematics of the western Mediterranean, in Coward, M.P., Dietrich, D., and Park, R.G., eds., Alpine Tectonics, Volume 45: London, Geological Society of London, p. 265-283.

Drury, S.A., 2001, Image Interpretation In Geology, 3rd Edition: Malden, MA (USA), Blackwell Science Inc. , 290 p.

El Alami, S.O., Tadili, B.A., Ait Brahim, R.L., and Mouayn, I., 2004, Seismicity of Morocco for the Period 1987–1994: Pure and Applied Geophysics, v. 161, p. 969-982.

Ensslin, R., 1992, Cretaceous synsedimentary tectonics in the Atlas system of Central Morocco: Geologishe Rundschau, v. 81, p. 91-104.

Evans, D., 1988, Multisensor classification of sedimentary rocks: Remote Sensing of Environment, v. 25, p. 129.

74 Faccenna, C., Piromallo, C., Crespo-Blanc, A., Jolivet, L., and Rossetti, F., 2004, Lateral slab deformation and the origin of the western Mediterranean arcs: Tectonics, v. 23, p. 21.

Fadil, A., Vernant, P., McClusky, S., Reilinger, R., Gomez, F., Ben Sari, D., Mourabit, T., Feifl, K., and Barazangi, M., 2006, Active Tectonics of the Western Mediterranean: Geodetic Evidence for Roll Back of a Delaminated Subcontinental Lithospheric Slab Beneath the Rif Mountains, Morocco: Geology, v. 34, p. 529.

Fedan, B., 1989, Evolution geodynamique d'un bassin intraplaque sur decrochements; le Moyen Atlas (Maroc) durant le Meso-Cenozoique. Geodynamic evolution of an intraplate basin over strike-slip faults; the Middle Atlas, Morocco, during the Meso-Cenozoic: Travaux de l'Institut Scientifique, Serie Geologie et Geographie Physique, v. 18, p. 142.

Fujisada, H., 1995, Design and performance of ASTER instrument: Proceedings of SPIE, the International Society for Optical Engineering, v. 2583, p. 16-25.

Gomez, F., Allmendinger, R., Barazangi, M., Er-Raji, A., and Dahmani, M., 1998, Crustal shortening and vertical strain partitioning in the Middle Atlas Mountains of Morocco: Tectonics, v. 17, p. 520-533.

Gomez, F., Barazangi, M., and Bensaid, M., 1996, Active tectonism in the intracontinental Middle Atlas Mountains of Morocco: Simultaneous crustal shortening and extension: Journal of the Geological Society (London), v. 153, p. 389-402.

Gomez, F., Beauchamp, W., and Barazangi, M., 2000, Role of the Atlas Mountains (northwest Africa) within the African-Eurasian plate- boundary zone: Geology, v. 28, p. 775-778.

Görler, K., Helmdach, F.-F., Gaemers, P., Heißig, K., Hinsch, W., Mädler, K., Schwarzhans, W., and Zucht, M., 1988, The uplift of the central High Atlas as deduced from neogene continental sediments of the Ouarzazate province, Morocco, The Atlas System of Morocco, p. 359- 404.

Gueguen, E., Doglioni, C., and Fernandez, M., 1998, On the post-25 Ma geodynamic evolution of the western Mediterranean: Tectonophysics, v. 298, p. 259-269.

Hanks, T., 2000, The age of scarplike landforms from diffusion-equation analysis: AGU Reference Shelf, p. 313-338.

75 Hanks, T.C., Bucknam, R.C., Lajoie, K.R., and Wallace, R.E., 1984, Modification of wave-cut and faulting-controlled landforms: Journal of Geophysical Research, v. 89, p. 5771-5790.

Harmand, C., and Cantagrel, J.M., 1984, Le volcanisme alcalin Tertiaire et Quaternaire du Moyen Atlas (Maroc): chronologie K/Ar et cadre geodynamique: Journal of African Earth Sciences, v. 2, p. 51-55.

Harmand, C., and Moukadiri, A., 1986, Synchronisme entre tectonique compressive et volcanisme alcalin: exemple de la province quaternaire du Moyen Atlas (Maroc): Bulletin de la Societe geologique de France, v. (8) II, p. 595-603.

Kanes, W.H., Saadi, M., Ehrlich, E., and Alem, A., 1973, Moroccan crustal response to continental drift: Science, v. 180, p. 950-952.

Lonergan, L., and White, N., 1997, Origin of the Betic-Rif mountain belt: Tectonics, v. 16, p. 504-522.

Lorenz, J., 1976, Triassic sediments and basin structure of the Kerrouchen Basin, central Morocco: Journal of Sedimentary Petrology, v. 46, p. 897-905.

Marrett, R., and Allmendinger, R., 1990, Kinematic analysis of fault-slip data: Journal of Structural Geology, v. 12, p. 973-986.

Martin, J., 1981, Le Moyen Atlas central, etude geomorphologique. The central Middle Atlas; geomorphologic study: Notes et Memoires du Service Geologique, v. 258 bis, p. 445.

Medina, F., and Cherkaoui, T.-E., 1991, Focal Mechanisms of the Atlas earthquakes, and tectonic implications: Geologische Rundschau, v. 80, p. 639-648.

Morel, J.-L., 1989, Etats de contrainte et cinématique de la chaîne rifaine (Maroc) du Tortonien à l'actuel: Geodinamica Acta, v. 3, p. 283-294.

Morel, J.-L., Zouine, E.M., and Poisson, A., 1993, Relations entre la subsidence des bassins moulouyens et la création des reliefs atlasiques (Maroc): un exemple d'inversion tectonique depuis le Néogène: Bulletin de la Societé Geologique du France, v. 164, p. 79- 91.

Pena, S.A., and G., A.M., 2006, Orbital remote sensing for geological mapping in southern Tunisia: Implication for oil and gas exploration: Journal of African Earth Sciences, v. 44, p. 203-219.

76 Platt, J.P., Soto, J.I., Whitehouse, M.J., Hunford, A.J., and Kelley, S.P., 1998, Thermal evolution, rate of exhumation, and tectonic significance of metamorphic rocks from the floor of the Alboran extensional basin, western Mediterranean: Tectonics, v. 17, p. 671-689.

Platt, J.P., and Vissers, R.L.M., 1989, Extensional collapse of thickened continental lithosphere: A working hypothesis for the Alboran Sea and Gibraltar arc: Geology, v. 17, p. 540-543.

Seber, D., Barazangi, M., Ibenbrahim, A., and Demnati, A., 1996, Geophysical evidence for lithospheric delamination beneath the Alboran Sea and Rif-Betic mountains: Nature, v. 379, p. 785-790.

Suppe, J., 1983, Geometry and kinematics of fault-bend folding: American Journal of Science, v. 283, p. 684.

Wells, D.L., and Coppersmith, K.J., 1994, New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement.: Bulletin of the Seismological Society of America, v. 84, p. 974-1002.

Ziegler, P.A., 1988, Evolution of the Arctic-North Atlantic and the Western Tethys, AAPG Memoir 43: Tulsa, OK (USA), American Association of Petroleum Geologists, 198 p.