Rupturing Continental Lithosphere (RCL)

RCL Continued...

6.2 Focus Site 2: Central and Northern Red Sea/Gulf of Suez

6.2.1 Introduction

The Red Sea and Gulf of Aden (Figure 7) are the closest modern analogs to the rifting and rupturing of continental lithosphere which formed the vast majority of “Atlantic-type” continental margins, and are the location where the processes that shaped the early development of rifted continental margins can be studied with the fewest tectonic complications. Nearly all of the passive continental margins of the Atlantic, Indian and Arctic Oceans were formed by the nucleation of an oceanic spreading center within a continental rift following an extended period of rifting. This process is presently occurring in the Gulf of Aden and Red Sea. The Gulf of Aden and Red Sea spreading centers developed within continental rift valleys cutting for 3500 km through the interior of a once contiguous craton (Beydoun, 1970; 1981; Stoeser and Camp, 1985) following a 15-30 m.y. period of continental rifting (Purser and Hötzl, 1988; Davison et al., 1994; Fantozzi, 1996; Omar and Steckler, 1995).

The opening in both rifts has been primarily extensional rather than transtensional (Joffe and Garfunkel, 1987; Jestin et al., 1994; Cochran, 1981; 1982; Colletta et al., 1988). Other than a single episode of dike emplacement dated at ~22 Ma (Bartov et al., 1980; Eyal et al., 1981), there has been virtually no volcanic activity within the northern Red sea rift and extension in the northern Red Sea has been accommodated primarily by rotation of large crustal fault blocks. It has only been recently with the establishment of localized centers of intrusion (deeps) that significant volcanic activity has occurred within the northern Red Sea Rift (Martinez and Cochran, 1988). In the very northern Red Sea, the deeps are localized centers of intrusion within the continental rift (Pautot et al., 1984; Cochran et al., 1991) while further south they have evolved into well developed seafloor spreading cells actively lengthening and propagating toward each other (Pautot, 1983; Bonatti et al., 1984; Cochran, 1983). The northern Red Sea is thus a non-volcanic continental margin at which both seafloor spreading has recently begun and the mid-ocean ridge is in the process of becoming established.

Thus, the Gulf of Suez and northern/central Red Sea allow the development of a non-volcanic continental margin to be observed and studied from late-stage continental rifting through the establishment of a mid-ocean ridge system to a well-developed young continental margin and ocean basin in a tectonic setting similar to that which has produced the majority of such margins throughout the geologic record. Furthermore, it is important to restate the need for correlating unequivocal “strain markers” across the Gulf of Suez and the northern/central Red Sea rift systems, thereby providing absolute constraints on the amount, transport direction, and location of total strain across the rifts (e.g., Sultan et al., 1993).

On the other hand, the rifting process in the southern Red Sea and westernmost Gulf of Aden has been dominated by the presence of the Afar hot spot (Mohr, 1970; 1978; Schilling, 1973; Morton and Black, 1975). As a result, rifting in the southern Red Sea and westernmost Gulf of Aden has been accompanied by copious volcanism (Coleman et al., 1975; 1983). In particular the southern Red Sea (south of ~22°N) is a volcanic margin where extension prior to the establishment of a localized oceanic spreading center at about 5 Ma was primarily accomplished through large-scale intrusion of new volcanic material (Gettings et al., 1986; Bohannon, 1986). Both the Red Sea and Gulf of Aden spreading centers are presently propagating into Afar (as the Erta Ale and Asal rifts respectively) (Mohr, 1970; Courtillot, 1980; Manighetti et al., 1998; Audin, 1999). The southern Red Sea thus presents the single best opportunity to study a recently formed volcanic continental margin and the Afar area presents an unparalleled opportunity to study hot spot - rift interactions.

The following summary of the Gulf of Suez is from Bosworth and McClay (2001). The structural and stratigraphic development of the Gulf of Suez reflects the interplay of five principal factors: 1) The presence of preexisting fault systems, penetrative fabrics and basement terrane boundaries, 2) eustatic variations, 3) changes in basin connectivity to the Mediterranean Sea and Indian Ocean, 4) rapid changes in African intra-plate stress fields, and 5) activation of the Levant-Aqaba transform plate boundary. The Gulf of Suez rift initiated in the late Oligocene as a result of the northeastward separation of the Arabia from the African Plate. North of Suez, extension is more diffuse but mostly focused on the Manzala rift buried beneath the Nile Delta. Microearthquakes and teleseismic events attest to continuing extension along major normal faults in the Gulf of Suez, Gulf of Aqaba and northern Red Sea regions.

The Gulf of Suez is the northern termination of the Gulf of Aden/Red Sea rift system, and the earliest syn-rift sediments are predominantly continental red beds with minor basalts. By the earliest Miocene, a shallow to marginal marine environment prevailed for most of the rift. The late early Miocene highstand allowed marine waters to mix between the Red Sea and the Mediterranean. Thick halite deposits formed in the late Miocene, and later sediment loading resulted in the formation of salt diapirs and salt walls.

Analysis of fault geometries, fault kinematics and syn-rift stratal relationships indicate that rift-normal extension predominated throughout the Oligocene to early middle Miocene development of the rift. In the middle Miocene, the Levant-Gulf of Aqaba transform boundary was established, linking the Red Sea extensional plate boundary to the convergent Bitlis-Zagros compressional plate boundary (Figure 7).

The northwest-trending Gulf of Suez is about 300 km long and the complete basin width, including onshore border fault systems, varies from 50 km at its northern end to ~90km at its southern end where it merges with the Red Sea (Figure 8). The rift is characterized by a zig-zag fault pattern, composed of north-south to north northeast-south southwest, east-west and northwest-southeast striking extensional fault systems both at the rift borders and within the rift basin proper. The fact that syn-rift sediments crop out on the both the Sinai Peninsula and western Egyptian desert implies that the focus of rifting likely propagated towards the center of the Suez rift, allowing flexural rebound engendered by later extensional faulting to expose the earlier syn-rift sediments (Figure 8).

There are three distinct rift segments or depocenters within the overall Gulf of Suez; the Darag basin at the northern end, the central basin or Belayim Province, and the southern Amal-Zeit Province (Figure 8). Each sub-basin is asymmetric, bounded on one side by a major northwest-trending border fault system with large throws (4-6 km in general) together with a dominant stratal dip direction toward the border fault system. Structurally complex accommodation zones, oblique to the rift trend, separate the three depocenters. The accommodation zones appear to be wide (up to 20 km) areas of complexly faulted blocks of variable dips and interlocking “flip-flop” conjugate fault systems. Within each of the three main half-graben there are second-order sub-basins formed by individual fault blocks, each of which has its own characteristic syn-rift stratigraphy.

6.2.2 Selecting the central-northern Red Sea/Gulf of Suez focus site

There are three areas of active rifting that have been foci for recent studies dealing with the development of an oceanic spreading center within a continental realm in addition to being viable locations for a focussed study of the rupturing of continental lithosphere; the Gulf of California/Salton Trough, the Woodlark Basin, and the Red Sea/Gulf of Suez/Gulf of Aden region. Each of these geographic areas represents a very different setting for continental rifting.

The Gulf of California developed by oblique extension in a primarily strike-slip regime (e.g., Larson et al., 1968; Larson et al., 1972; Moore and Buffingham, 1968; Lonsdale, 1989) just landward of the continental margin. Slicing a thin sliver off of a continent may be a relatively common phenomenon (Steckler and ten Brink, 1986) and represents a mechanism for the creation of exotic terranes (e.g., Coney et al., 1980; Churkin et al., 1982). However, the Gulf of California is a very different tectonic regime compared with the rift settings responsible for the majority of “Atlantic-type” continental margins.
Woodlark Basin rifting appears to have originated in response to a change in stress regime related to plate convergence and subduction of the Australian plate (Weissel et al., 1982; Benes et al., 1994). The present Woodlark Basin spreading center extends from the Simbo transform (which extends to the Solomon Trench; Crook and Taylor, 1994) to the Papuan Peninsula of New Guinea, into which the rift is propagating (Crook and Taylor,1994; Taylor et al., 1995). The time period over which the continental rifting stage developed prior to the establishment of the oceanic spreading center is rather short (< 4 m.y., Taylor et al., 1995) compared to the rifting phases at most passive continental margins that typically last for 15 m.y. or longer. Recent work has shown that the bathymetry and structure of the western Woodlark Basin has been significantly shaped by the thermal consequences of subduction processes (Fang et al., 1997), especially the addition of subduction-related fluids that, in turn, have modified significantly the rheology of the fore-arc and back-arc crusts. The Woodlark Basin thus represents a special and unusual tectonic setting and the extent to which observations from the Woodlark Basin can be used to understand the development and structure of other margins is not clear.

Only the Red Sea and Gulf of Aden present an opportunity to observe and study the initiation of a young rifted continental margin that formed by the deformation of stable cratonic continental crust. As such, the Red Sea and Gulf of Aden represent rift settings characteristic of the vast majority of continental margins around the world. During the Snowbird meeting (January, 2000) in which the community deliberated on the RCL focus sites, the Red Sea and the Afar/Gulf of Aden were kept separate because of the nature of the science considered important in these two areas. Following the unanimous choice of orogenic rifting in the Gulf of California by the Snowbird meeting participants, the cratonic rifting choice become the central/northern Red Sea region, with the Afar/Gulf of Aden region being rejected based primarily on logistical problems of working onshore in Somalia and Aden, the political instability of Aden, and the then-war between Eritrea and Ethiopia.§ Since the Sharm el-Sheik meeting, it has become clear that organized seafloor spreading is indeed occurring to the north of the Eritrean border. To conform with the original Snowbird 2000 meeting intent of the central/northern Red Sea study region, Eritrea is not part of this region. If, and only if, it can be demonstrated that rifting has not progressed to seafloor spreading in the Sudanese/Saudi section of the Red Sea is there a need to expand the research region to include northern Eritrea and the Sudan.

6.2.3 Geophysical data available from the central-northern Red Sea

Geophysical surveys of the central and northern Red Sea have been conducted by the oil industry and also by a number of academic institutions. A relatively large geophysical database has been collected by various institutions over the last 25 years. The total amount of digital data presently in our data base amounts to 243,418 depth, 178,980 gravity, and 155,544 magnetic measurements.

The largest blocks of geophysical data were collected during multichannel seismic surveys of the Egyptian side of the Red Sea from 25°N to 27°40’N carried out by Exxon and Phillips Petroleum in the mid to late 1970s. These surveys consisted of dip lines at 2- 4 km spacings nearshore and ~8 km spacing away from the coast with strike lines at 5-8 km spacing. We have obtained the bathymetry, gravity and magnetics data from the Exxon and Phillips surveys and have permission from the Egyptian government to publish it. Some Exxon data are also available, but only in analog form. A few additional Exxon seismic lines were published by Miller and Barakat (1988).

A similar survey was carried out on the Saudi Arabian side north of 24°N by Preussag AG for the Saudi government in the early 1980s. As Preussag has closed its marine survey division, inquiries have indicated that the data no longer exist in digital form in Germany. It is unclear whether it exists in Saudi Arabia. The German BGR has released copies of detailed bathymetric and magnetics maps, as well as maps of depth to various seismic horizons. These maps are at scales ranging from 1:100,000 to 1:250,000. Degraded “simplified” versions of some of these maps have been published by Richter et al., (1991).
In addition to the industry data, data from surveys carried out by U.S (LDEO and WHOI), French (IFREMER and Ecole Normale Supérieure), British, Italian, Russian and Israeli academic institutions as well as bathymetry data from about 20 British and U.S. Navy transits of the northern Red Sea are also available. The two most extensive academic surveys were those carried out by LDEO on Conrad and by IFREMER on Charcot. Track charts for these individual cruises as well as the details of data acquisition can be found in Martinez and Cochran (1988) and Guennoc et al., (1988). Both the Conrad and Charcot surveys collected 24 beam SeaBeam and single-channel seismic reflection data.

Seismic refraction lines across the region were collected between 1978 and 1986 by two research groups. A German group from Hamburg carried out several experiments using”MARS-66" portable seismometers and explosives for on-shore lines in Egypt and Saudi Arabia and OBSs and either explosives or airguns for marine lines. These profiles are summarized by Rihm et al., (1991) and discussed in more detail in the “grey” literature (Makris et al., 1979; 1983) and in University of Hamburg theses. A series of expanding spread profiles were carried out by the French École Normale Supérieure off Egypt in 1986. This work was reported by Gaulier et al., (1988). The seismic data shows fairly typical crustal thicknesses of 30-40 km onshore away from the Red Sea. The crustal thickness decreases slightly to 20-30 km within the onshore portion of the rift and then abruptly decreases to less than 10 km seaward of the coast. Thermal gradients, conductivities and heat flow values were published in Martinez and Cochran (1989), which, when combined with onshore measurements in Egypt published by Morgan et al., (1985), give a good picture of the variation in heat flow from the unrifted craton completely across the rift.

The majority of the available geophysical data from the northern Red Sea was obtained between 1975 and 1990. The data is well located using a combination of satellite and LORAN navigation and the bathymetry, gravity and magnetics data are generally of very good quality. As mentioned earlier, several cruises in the mid and late 1980s included SeaBeam data. However, in all cases this was the older SeaBeam system. As a result of the shallow water depths, the swath width was generally less than 1 km and the SeaBeam data were of limited value for general surveying although it was very useful for surveys of specific features such as deeps (Pautot, 1983; Pautot et al., 1984; Cochran et al., 1986). However, the density and generally high quality of the available data has allowed an understanding to be obtained of the structure of the northern Red Sea and is more than adequate to use in planning further experiments using the more sophisticated geophysical instrumentation now available.

6.2.4 The “Salt Problem” in the central-northern Red Sea

The main exception to the conclusion of the previous paragraph deals with the quality of multichannel seismic reflection data. The Exxon seismic sections published by Miller and Barakat (1988) show the base of the Miocene evaporites in many places but very little structure below the evaporites. This has led to a general opinion that the thick Miocene evaporites prevent imaging of deeper structure and basement within the Red Sea.
In reality, this is not quite accurate. Izzeldin (1987) presents a seismic section off of Sudan in which he is able to image a reflector, interpreted as basement, through the salt to a depth of 2.9 sec (two-way travel time). Also Preussag was able to prepare maps of the depth to a number of sub-salt horizons along the Saudi coast (see summary chart in Richter et al., 1991).

Much of the difficulty with MCS data from the Red Sea is that all of the available or published data was acquired 20-25 years ago using what is now very antiquated equipment. The Exxon survey was carried out with a 48-channel, 2350 m-long streamer, while the Preussag Saudi survey utilized 24-channel, 2400m and 30-channel, 3000 m streamers with a DFS V recording unit.

6.2.5 Geologic background of the Red Sea

The Red Sea occupies a 2000-km long rift bounded depression (Figure 7). The rift shoulders average between 1000 m and 3000 m in elevation and expose a variety of Pan- African (late Proterozoic) granitic, metamorphic and mafic igneous rocks (Shackleton et al., 1980; Stern, 1984). Continental rifting appears to have begun nearly simultaneously along the entire length of the Red Sea in the Oligocene at ~ 34 Ma with the main phase of extension beginning at about 22 Ma (Omar and Steckler, 1995).

The morphology of the Red Sea consists of narrow marginal shelves and coastal plains, and a broad “main trough” with depths of about 400-1100 m. In the southern Red Sea, the main trough is bisected by an axial trough ~60 km wide with depths of up to 2000 m. A mid-ocean ridge spreading center with well developed seafloor spreading magnetic anomalies occupies the axial trough from about 15°N to 19°30’N (Roeser, 1975; Cochran, 1983; Miller et al., 1985; Garfunkel et al., 1987). Spreading appears to have nucleated at about 17°N at ~ 5 Ma and to have propagated north and south from there (Roeser, 1975; Courtillot, 1982; Cochran, 1983). The spreading center becomes discontinuous north of 19°30’N and passes into a transition zone made up of a series of isolated seafloor spreading cells (Cochran, 1983; Pautot, 1983; Bonatti, 1985; Bicknell et al., 1986). These “deeps” are separated by “intertrough zones” which are shallower, broader, and covered with highly faulted sediments including both the Miocene evaporites and post-Miocene pelagic sediments (Searle and Ross, 1975; Izzeldin, 1989).

An organized mid-ocean ridge is not observed in the northern 500 km of the Red Sea (Figure 9). The region is characterized by terraced bathymetry stepping down to an axis of deep water which is characterized by faulted and deformed sediments (Martinez and Cochran, 1988). Small deeps, generally associated with large dipolar magnetic anomalies are spaced along the axial depression (Pautot et al., 1986; Guennoc et al., 1988; Cochran et al., 1986). These deeps are much smaller than the central Red Sea transition zone deeps and usually are floored by sediments, although a volcanic peak is found in Shaban Deep near 26 10’N (Pautot et al., 1984).

6.2.6 Structure of the Red Sea

The central-northern Red Sea can be divided into two distinct regions (Figure 9), a main trough often referred to as “marginal areas” although they actually occupy most of basin, and a 15-30 km-wide “axial depression” (Martinez and Cochran, 1988; Cochran and Martinez, 1988). The bathymetry of the marginal areas forms a series of terraces 20-30 km wide, generally at depths of about 600 m, 800 m, and 950 m. These terraces are separated by steeper slopes or escarpments which often appear to be fault controlled.

Free-air gravity anomalies form a pattern of elongate high-amplitude (50 mGal) highs and lows which are oriented subparallel to the trend of the rift and extend for 50-70 km along strike. The free-air gravity anomaly highs are persistently located on the seaward edges of the bathymetric terraces Gravity contours are systematically terminated or offset across NE-SW trending zones. Bathymetric contours often are also offset at locations where the gravity contours are offset or terminate (Martinez and Cochran, 1988; Cochran et al., 1991). The gravity anomalies display the pattern of basement relief, interpreted as a series of tilted fault blocks 15-30 km across and roughly 60 km in length separated by accommodation zones which absorb the differential motion between adjacent sets of fault blocks (Martinez and Cochran, 1988).

A region of deep water 1100-1300 m deep and from 15 to 30 km wide referred to as the “axial depression” (Cochran et al., 1986) is a consistent feature of the northern Red Sea. The axial depression differs from the oceanic “axial trough” of the southern Red Sea in that sedimentary sequences, including both the Miocene evaporites and the post-Miocene pelagics, are continuous across it, lineated magnetic anomalies are not present and it is shallower. The axial depression often appears to be fault bounded, particularly in the vicinity of deeps. It is marked by a free-air gravity minimum with a relative amplitude of 30-60 mGal and is the location of the maximum heat flow on each of our heat flow traverses (Martinez and Cochran, 1989). Deformation of the sediments in the axial depression is more intense and concentrated than in the marginal areas and frequently extends to the seafloor. Martínez and Cochran (1988; 1989) argue that both the distribution of faulting and numerical modeling of heat flow data require that deformation previously occurred over a wide area, but has relatively recently become focused in the axial depression.

The axial depression is not only the locus of recent deformation, but is also the location of a series of axial deeps spaced at 50–75 km intervals along it. These are all small northern Red Sea type deeps (Pautot et al., 1984; 1986; Cochran et al., 1986; Guennoc et al., 1988). In the area north of 26°N, where accommodation zones have been identified, the deeps are located almost exactly at the midpoint of segments, halfway between the accommodation zones. Except for the northernmost deep, near 27°20N, 34°20’E, the deeps are associated with large normally magnetized, dipolar magnetic anomalies which appear to result from recent, localized intrusions (Cochran et al., 1986). In the majority of cases, there is not a single anomaly over the deeps, but rather a pair of dipolar anomalies. A detailed study of Conrad Deep showed that the bodies responsible for the magnetic anomalies are not located under the deep itself, but rather are beneath the base of the fault scarps on either side of the deep and are elongated parallel to the fault scarps (Cochran et al., 1986), which strongly suggests that the faults bounding the axial depression have been used for the ascent of the magma.

It appears on both bathymetry and gravity maps that the axial depression is not simply a continuous axis of deep water as suggested in published descriptions (Pautot et al., 1984; Guennoc et al., 1988; Martinez and Cochran, 1988; Cochran et al., 1991), but is systematically segmented (Figure 9). Bathymetric depth and the amplitude of the gravity lows both decrease away from the deeps with minima at the accommodation zones identified by Martinez and Cochran (1988). The accommodation zones are not simply saddle points, but also offset the axial depression. The bathymetric and gravity lows associated with deeps do not intersect, but rather overlap without joining. This can be seen, for example, in Figure 9 where the gravity lows centered at Conrad Deep (27°03’N) and the 26°36’N deep curve away from each other as they approach the accommodation zone between the two deeps. The two gravity minima are separated by a 10 km-wide, 15 mGal relative high. The axial depression thus appears to be divided into discrete, independent segments separated by the same accommodation zones that define the geometry of the basement fault blocks within the marginal areas. Within each segment, there is an axial deep located almost exactly half way between the accommodation zones and associated with high-amplitude normally-magnetized dipolar magnetic anomalies which appear to result from large recent intrusions (Pautot et al., 1984; Cochran et al., 1986).

Models have been developed for the establishment of an oceanic spreading center in a continental rift based on the present geophysical observations (Martinez and Cochran, 1988). It is hypothesized that the Red Sea rift began in the Oligocene as a series of linked half graben as presently observed in the East African Rift (Bosworth, 1985; 1994; Ebinger et al., 1987; Ebinger, 1989). By the mid-Miocene, the initial half-graben had evolved into sets of rotated fault blocks as presently observed in the Gulf of Suez (Garfunkel and Bartov, 1977; Colletta et al., 1988; Bosworth, 1994). In this stage, referred to by Bosworth (1995) as a “high-strain rift,” deformation and subsidence became more focused along the basin axis, with extension occurring along a new system of higher angle, planar faults (Bosworth, 1994; 1995). The Gulf of Suez rift is still segmented by accommodation zones and all of the blocks in a segment have a consistent sense of dip (Colletta et al., 1988; Bosworth, 1994).

An additional 100 km of extension has occurred in the northern Red Sea since the Gulf of Suez was cut off in the mid-Miocene. Buck et al., (1988) and Martínez and Cochran (1989) have argued the heat flow in the northern Red Sea requires that extension (with thinning of the crust and lithosphere) was spread throughout the rift but has recently become focused near the axis. This is consistent with the model developed by Bosworth (1994; 1995) for the Gulf of Suez. It is also consistent with the more symmetric appearance of the northern Red Sea and with the observation of more intense sediment deformation in the axial depression (Martinez and Cochran, 1988). It appears that in this process, the basic tectonic framework of the rift has remained sets of rotated fault blocks separated by accommodation zones.

Thermal modeling (Buck et al., 1988; Martinez and Cochran, 1989) also suggests that as the extension becomes concentrated, some degree of partial melt will be generated beneath the axis. This melt appears to have reached the surface at the deeps which are characterized by large amplitude dipolar magnetic anomalies. Most of the deeps do not have a single magnetic anomaly but rather pairs of anomalies centered over each of the faults bounding the axial depression implying that the magma was intruded along the faults (Cochran et al., 1986). We hypothesize that the deeps develop into small sea-floor spreading cells which propagate and grow together to form an oceanic spreading center. The segmentation of the newly formed mid-ocean ridge is thus hypothesized to be inherited from the geometry established during continental rifting.

6.3 Allied field studies: Studies of deeply-exhumed extended terrains

Deeply-exhumed extended terrains are important for a full understanding of the processes of lithospheric rupture and extension. Such terrains, where well exposed, can offer an accessible 3-D geometrical picture of fossil fault systems and shear zones, rock and mineral fabrics, metamorphism or melting, and alteration down to very small scales. They can be sampled in whatever level of detail is necessary for each technique and problem. Similar information at depth in active systems is hard to obtain in such detail, because the rocks can only be accessed in very limited locations (i.e. occasional boreholes). Therefore, deeply exhumed extended terrains are important locations for addressing certain key questions of rupturing of the continental lithosphere.

The questions that can be addressed in these terrains will vary according to the rock types and their preservation or subsequent alteration, but ideally will include some or all of the following: unroofing history determined by geochronology, geothermometry, petrology (if melts were formed) and geobarometry; extensional deformation history from mineral fabrics, shear sense indicators, fractures, and veins; the nature of the crust and mantle involved in the deformation (from compositions of basement and/or syntectonic igneous rocks); the nature of any previous deformation that may be unrelated to the later extensional deformation that has affected the region; and the lithospheric-scale geometry of extensional systems.

One question that remains in some cases of exhumed extensional fault systems is, how is the deformation in a “metamorphic core complex” (i.e. the mylonite zone) actually related to the extension and the low-angle fault that unroofed the system? Similarly, in at least one case where the ocean-continent transition is exhumed, it is thought to be a zone of continental mantle (see summary by Whitmarsh et al., 2001). This appears to differ from what is present in either of our focus sites; yet, detailed study of this or other regions of transitional crust will help to establish constraints on composition, fabric, and timing of deformation. These can be compared with the focus areas to yield a fuller understanding of the range of possible behaviors as continental lithosphere ruptures to form an ocean basin.

7. Critical Field, Laboratory, and Experimental Efforts

7.1. Active-source seismology

Images and seismic velocities are obtainable at scales useful for probing the RCL Focus Sites to depths of about 50 km using newer experimental facilities and focused observational programs. Below the crust, depth and resolution are limited by the difficulty in propagating energy to great depths, and passive-seismic methods become the more powerful tools.

Conventional industry-style exploration reflection seismology will be essential for imaging the stratigraphic response to lithosphere rupturing, and for mapping the threedimensional patterns of faulting in the brittle crust. The basin subsidence history, preserved in the basin-filling stratigraphic sequences, is the most sensitive record of lithospheric extension and thermal evolution during rupturing. Significant numbers of 2D profiles already exist for both Focus Sites, but some 3D data volumes may be necessary to properly understand areas of complex fault interactions, particularly perhaps those between strike-slip and normal faults, or between faults and active magmatic intrusion. Should it prove possible to identify, and appropriate to drill into, an active low-angle normal fault, 3D seismic data will be required before drilling. Some 2D datasets should be available from the hydrocarbon industry for the Red Sea focus site, and other data-sets have been and must continue to be acquired with NSF funding in both focus sites. For example, a 2D “high-resolution” profile in the northern Gulf of California shows well the complex stratigraphic and faulting record preserved in the Delfin Basin (Figure 4), and other profiles in the Gulf of California seem to show uppercrustal intrusions (Gonzalez-Fernandez, 2000, MARGINS abstracts). Reflection imaging will also be important to study the lower crust, since prominent, laminated lower-crustal reflectivity is believed to be a signature of extensional strain and magmatism (e.g., Reston, 1990; Warner, 1998).

Reflection and refraction techniques become more powerful when combined. Closely spaced seismographs (or ocean-bottom seismometers/hydrophones (OBS/H) offshore) along modern normal-incidence reflection lines have been used to extend structural imaging to depth as well as provide unique velocity data. Tomographic analysis of these data also provides the crustal and upper-mantle velocity field to combine with the structural analysis possible from near-vertical reflections. Velocity measurement is the best way to identify mafic intrusions (“rift pillows”) formed at the base of the crust in some active rifts, including the Salton Trough (Fuis et al., 1984).

7.2. Earthquake-source seismology

The greater energy release and global distribution of earthquake sources allows their signals to be used to extend both structural and velocity mapping beneath the crust. Most analogous to the refraction profiles of active-source seismology are the 3D velocity tomograms of the mantle (some from surface wave analysis, others from body-waves), which show unusually low upper-mantle velocities, ascribed to unusually high upper mantle temperatures, below both western North America (Gulf of California, Basin and Range province) and the Red Sea/East African rift system (e.g., Dziewonski, 2000). Most analogous to the normal-incidence multichannel seismic sections are the receiver functions that extract P-to-S and S-to-P conversions formed at the 660-km, 410-km and Moho discontinuities, or other relatively abrupt velocity changes. Dense arrays of broadband seismographs deployed for months or years now collect data from enough sources to allow stacking and migration of receiver functions, with the same increase in resolving power that multi-channel seismic data shows over single-channel seismics. Preliminary receiver-function studies have already been carried out to map Moho depths in part of the northern Gulf of California (Lewis et al., 2000; in press), but more such work will surely prove necessary in both Focus Sites.
Other passive seismic techniques provide information that is very different from active-source techniques. Careful hypocentral location using local seismograph networks can delineate active faults and zones of active magmatism in both the crust and mantle (e.g., Rebollar, 2000; MARGINS abstract). Passive seismic arrays using portable PASSCAL-type deployments are rapidly expanding our understanding of the Earth’s deep interior in many ways. Attenuation studies can provide constraints on temperature variations beneath rifts independent of those provided by velocity tomography or heatflow and xenolith studies.

Shear-wave splitting measurements can determine the anisotropy of the upper mantle and crust (Savage and Sheehan, 2000 for Basin and Range Province). The splitting observations constrain the flow-induced fabric of the mantle to the extent that olivine crystals align and produce bulk anisotropy, and provide tests of dynamical models of rifting. Because this anisotropy typically includes components of both modern asthenospheric flow and fossil lithospheric anisotropy, passive seismic deployments must be broad enough to capture the areal variation of flow, and long-term enough to record sufficient suitable earthquakes at a range of back-azimuths to enable resolution of multi-layer anisotropy, if present. A seismic-related project is already funded to carry out an appropriate teleseismic study around the Gulf of Suez; despite the recent establishment of the NARS-Baja seismic array (Trampert et al., 2003), additional portable instrument deployments will be necessary for more detailed studies in the Gulf of California. Although mantle anisotropy is now routinely determined by passive arrays, determination of crustal anisotropy is much more difficult because the layer thickness in which such anisotropy is developed is much smaller (c. 10 km) than in the upper mantle (c. 100 km). Yet knowledge of lower-crustal anisotropy may yet prove the best determinant of the way the weak lower crust is responding to the boundary forces that ultimately control rifting.

7.3. Geodesy

Space geodesy (GPS, DORIS, VLBI, SLR) gives us a present-day “snapshot” of plate motion and lithospheric deformation, and has the potential to contribute to our understanding of rift initiation and evolution in several ways.

First, space geodesy can provide information on kinematic boundary conditions: What are the motions of adjacent plates or blocks, and how might these motions help or hinder the rifting process? These boundary conditions may be crucial to determining the long term fate of a rift, in terms of eventual development of sea floor spreading and oceanic crust. Consider the modern Red Sea, formed by northward motion of the Arabian plate relative to the Nubian plate (Africa west of the East African Rift). Comparison of modern Arabia-Nubia motion (from space geodesy) to a several million year average (from sea floor magnetic anomalies or from geological models that uses these and other data) indicates that Red Sea spreading is slowing down. One explanation is that Arabia is colliding with the Eurasian plate to the north, and the resulting thickened crust and excess elevation of the Zagros and Caucasus Mountains exerts south-directed gravitational body forces on the Arabian plate that act to slow rifting in the Red Sea. Since the history of Red Sea opening is reasonably well described, one way to test this would be to investigate the history of crustal shortening in the fold and thrust belts on the northern boundary of the Arabian plate.

The mechanics and thermodyamics of ultra-slow sea floor spreading is another area of interest in rift development, since rifting of continental lithosphere may start out slowly, but presumably must accelerate to some minimum speed if sea floor spreading is to be achieved. By obtaining accurate measures of present day spreading rates for comparison with other geologic data, we may gain some useful insights. Using the Red Sea example, the GPS-determined angular velocity for Arabia-Nubia accurately defines the spreading rate increase from north to south. Sea floor spreading in the modern Red Sea seems to occur when the spreading rate reaches about 10 mm/a (full rate). At rates slower than this, isolated zones of magma upwelling occur, but do not coalesce into organized spreading centers.

The partitioning of strain in developing continental rifts is also of interest, in part because it may help us to understand how rifts develop. A current paradox is that the stresses available to break continental lithosphere seem to be too small, given current estimates of lithospheric strength. GPS in particular is well suited to defining in some detail the surface deformation field around active rifts. By comparing measured deformation to mechanical models of deforming lithosphere, we can gain insight into the deformation mechanisms. As GPS data become more precise, it is apparent that mechanical models of the lithosphere must go beyond the simple elastic half space formulation that are the mainstay of current crustal deformation models for geodetic data. Developing improved mechanical models will require constraints from heat flow, geological mapping, paleoseismic studies, and laboratory data on the strength of common crustal materials as a function of temperature and strain rate.

7.4. Geodynamic modeling

Geodynamic models are crucial to our understanding of the physical processes governing rifting of continental lithosphere and subsequent initiation of seafloor spreading. Models will provide the framework within which diverse onshore and offshore datasets may be synthesized to illuminate the intricacies of processes such as: strain localization and the evolution of fault systems; magma emplacement and how it interacts with deformation; hydrous fluid flow and sediment dynamics; strain partitioning during extensional and transtensional rifting; vertical partitioning of strain within the lithosphere and interactions between crust and mantle during continental rupture.

The ultimate goal of geodynamic models, however, will be to answer broad-scale questions regarding how continental extension in a localized zone of rifting leads to the initiation of seafloor spreading and thus to the birth of new plate margins. Fundamental to this goal is an assessment of the relative importance of plate-tectonic forces versus locally-derived body forces in driving the rupture of continental lithosphere. This question is intimately related to the stress conditions under which continental lithosphere ruptures and thus to the fundamental “low-strength” paradox in continental deformation (extreme examples of which include low-angle detachment faults). Additionally, in order to understand forces driving continental rupture, it is crucial to understand the extent to which previous tectonism “pre-conditions” the lithosphere and may enhance processes leading to rifting and the initiation of seafloor spreading.

Geodynamic models must, therefore, integrate surficial constraints on continental rupture processes (onshore and offshore geodetic strain rates, geologic slip estimates, strain partitioning as observed in the upper crust, conjugate margin geometry, fault segmentation, and magma emplacement) together with constraints on the threedimensional architecture and deformation of the plate margin (crust and mantle lithosphere thickness, seismic velocities, electrical conductivity, heatflow, petrology of crust and mantle xenoliths, inferred deformation fabrics from seismic anisotropy). Important issues to be addressed by geodynamic modeling include:

• The initiation and evolution of onshore fault systems during continental rifting and their relations to the offshore ridge-transform system following the transition to seafloor spreading
• Strain partitioning in the upper continental and oceanic crust during oblique transtension
• Strain partitioning within continental crust (decoupling of upper and lower crust) and the importance of low-angle detachment faulting
• Interactions between crustal extension, crust or mantle flow, and magma emplacement
• The relative importance of plate tectonic versus local body forces in governing continental rifting and the transition to seafloor spreading.

In both the Gulf of California/Salton Sea and the Red Sea/Gulf of Suez focus sites, the lateral variations in the transition to sea-floor spreading might be exploited in geodynamic models as a proxy for the temporal evolution of the process of continental rupture.

7.5. Magnetotellurics, heatflow, flexural strength and seismic attenuation

Magnetotellurics (MT) is an electromagnetic prospecting method in which orthogonal components of the horizontal electric and magnetic fields, induced by natural primary sources, are measured simultaneously as a function of frequency and are used to image the Earth’s electrical resistivity structure from depths of a few 100 meters to several 100 kilometers. The MT method is based on frequencies in the order of 10-3 Hz to 1 Hz. In general, rocks containing fluids such as water or rocks at high temperature will have a low resistivity. Similarly, dry and cold rocks will have high resistivities. In contrast, seismic Q values are a measure of seismic energy dissipation and relate to the presence of fluids and augmented by temperature variations. Enhanced electrical conductivity at subsolidus temperatures is principally caused by the presence of fluids. Consequently, conductive pathways serve as a geophysical tracer for the distribution of fluids and/or large temperature variations in the lower crust and mantle. Likewise, strain anisotropy may also affect both fluid permeability and electrical and seismic transmission pathways, thereby helping to determine the directions of preferential extension and plastic strain in the lower crust and mantle.

Perhaps the best proxy to help decipher the amplitude of involvement of the lithospheric mantle, especially when coupled with the patterns of regional subsidence, magnetotellurics and seismic attenuation, will be patterns and magnitudes of the surface heatflow. In general, high-quality heat flow, when coupled with measurements of sediment thermal conductivities, provides crucial information on the pre-deformed and deformed thermal structure of the continental lithosphere, which in term provides an important constraint or state variable for seismological, petrological and rheological models for the deforming lithosphere.

Plate flexure helps to redistribute the loads created by the extension process and therefore is a critical factor in controlling the regional geometry of extended lithosphere. In order to estimate the total extension of the upper crust (from, for example, the distribution of surface faulting as mapped from surface mapping or reflection seismics), the lithospheric strength as a function of space and time needs to be determined. As demonstrated by Karner et al., (2000), extension and lithospheric strength inversely interact to produce a given basin geometry; equivalent basin geometries are obtained for either low flexural strength and large extension or high flexural strength and small extension. The best way to constrain the flexural strength of the lithosphere is by forward modeling the free-air and Bouguer gravity of the basin system, both in terms of the basin architecture and the adjacent rift flank topography. Because gravity anomaly amplitudes are very sensitive to the mechanical or flexural strength of the lithosphere, it is thus imperative to collect high-resolution gravity across the Gulfs of California and Suez deformation zones and adjacent regions.

8. Databases and the hydrocarbon industry

Over the last few decades, there has been a steady but irreversible shift with respect to the ownership of extensive Earth Scientific data bases. Thirty to forty years ago, it was the academic institutions like Lamont and Scripps that were the custodians of worldwide data sets. These data consisted primarily of underway geophysical data such as single-channel seismics, gravity, magnetics, and bathymetry, in addition to dredge, core, and box-core samples. Nevertheless, these data represented important if not the only data available across many continental margins for subsequent hypothesis development or defining the “state of knowledge” of a region. However, the rapid growth of the oil industry over this same period, the drilling of onshore and offshore wells within virtually all of the world’s basins and margins, reconnaissance and detailed field mapping, geotechnical laboratory experiments and field sampling, reservoir characterization, remote sensing surveys, all coupled with the major advances in 2-D, 3-D and 4-D reflection seismic acquisition and processing, have resulted in the generation of immense geological and geophysical data bases that reside within the oil industry. The traditional funding organizations are not in a position to compete with industry in terms of data quality, coverage and rate of data acquisition, nor can the funding agencies afford to duplicate the acquisition of these data sets.

It is clear that MARGINS needs to facilitate collaboration between the oil industry and university researchers on problems of mutual interest. The MARGINS Office can play a prime role here by acting as a liaison between researchers and the oil industry in trying to facilitate the release of critical data sets. An excellent way of fostering this collaboration is via offering interested oil companies (non-voting) membership on the MARGINS or InterMARGINS Steering Committees.

9. National and International cooperative programs

A number of major NSF Earth Science initiatives and international MARGINS programs are currently being planned and will potentially impact the various MARGINS initiatives. For example, a major NSF-requested Major Research Equipment for the Earth Sciences (EAR), termed Earthscope, was recently funded in part by Congress. In contrast, the UK and European Earth scientific communities have organized themselves to solve first-order problems considered important for both academia and the oil industry.

9.1 Earthscope

Earthscope comprises four facility-based components; PBO, USArray, SAFOD and InSAR, all of which overlap in either technique or scientific endeavor with RCL objectives. The challenge is to take advantage of these “facility initiatives” during the execution of the MARGINS program. In order to highlight these overlaps, a summary of each of the Earthscope components are briefly summarized.

The goal of the USArray facility will be to routinely image the crust and mantle beneath the U.S. (and overlapping areas in Canada and Mexico), using an array of portable seismic instruments, thereby providing subsurface imaging of the plate boundary zone, obtaining high-quality constraints on the deformation processes; mechanical (fault zone), thermal (topography), mantle (anisotropy), and density (tomography and gravity).
The objective of the Plate Boundary Observatory (PBO) facility is to supply primarily geodetic data to characterize the 3D deformation (velocity) field of the western U.S., Alaska, and Baja California. It will thus define plate boundary dynamics and crustal rheology, define distribution and timing of the active tectonic processes and their relation relation to geology, and aid in understanding the physics of the earthquake process, and of volcanic processes.

The objective of the San Andreas fault observatory at depth (SAFOD) facility is to oversee drilling into the seismogenic portion of the San Andreas fault. The SAFOD mandate is to provide “hard” constraints on seismic, physical, and rheological properties of the San Andreas crust by drilling a 4 km-deep hole close to the hypocenter of the 1996 magnitude 6 Parkfield earthquake, where the San Andreas fault slips through a combination of small-to-moderate magnitude earthquakes and aseismic creep.

Although strictly a NASA initiative, the Synthetic aperture radar interferometry (InSAR) project will be used as a satellite tool to provide spatially continuous maps of the displacement field over the 100 km-swath width imaged by the satellite radar in those areas be investigated with USArray and PBO.

9.2 UK Ocean Margins LINK Program

The UK together with Ireland shares an ocean margin over 1500 km in length containing valuable oil and gas reserves but the economic benefits of this huge area have not yet been fully determined. UK MARGINS program, formally termed the Ocean Margins LINK program and funded by the Natural Environment Research Council (NERC) and the UK hydrocarbon industry, will focus on three main themes: 1) Deep structure and rifting processes; 2) Sedimentary processes, sediment movement and slope stability; and 3) Fluid flow particularly into and out of the seabed, including its relationship with and effect on deep-water faunas. The model presented by the LINK program is novel and one that the US MARGINS program should attempt to emulate.

In addition, the Ocean Margins LINK program is also designed to focus the research capabilities of the UK science base, working in partnership with industry on the challenges facing the industry in exploring for, and developing deep-water oil fields. Specifically, the Ocean Margins LINK program will contribute to the following industry “challenges”: 1) Improved prediction in exploration and reservoir characterization on the UK margin and globally; 2) Gas hydrates as a hazard and potential energy source; 3) Prediction of deepwater geohazards; and 4) Sustainable management of hydrocarbon resources and the deep-water environment.

The program is intended to stimulate industry and academic community partnership in key technological sectors and in doing so, enhance the competitiveness of UK industry and quality of life through the support of managed programs. LINK aims to accelerate commercial exploitation of science, engineering and technology. The program also helps stimulate industry to increase the resource of R&D within universities. A LINK project must have at least one university and one company, with access to data related to the project being granted by the collaborating companies.


EuroMARGINS was created to address fundamental scientific questions concerning the origin, structure and evolution of passive continental margins concentrating on the following thematic issues: Rifting processes, sedimentary processes and products, and sub-sea floor fluid flow systems. As such, EuroMARGINS shares a common goal with the US MARGINS program in many of its objectives. Both active and inactive systems plan to be studied, with the various North Sea and Mediterranean margins and basins being sites of focussed research. EuroMARGINS research necessarily needs to be of societal importance, especially as it relates to hydrocarbon exploration and production, the economic potential of gas hydrates, deep-sea biota and, the occurrence of natural hazards.

With funding through the European Science Foundation (which acts as a financial broker for monies paid by member European countries for MARGINS research), implementation of the EUROMARGINS program will be carried out in two main phases. Phase I will focus on two “target” areas that are of immediate priority for Europe—one on the Northwest European margin and the other in the Mediterranean Sea. The centerpiece of Phase I will be the first 3-D seismic tomography and imaging of the crust and upper mantle in the transitional region between the continent and ocean. The seismic studies will be accompanied by sonar studies of the mechanics and kinematics of slope instabilities; the dynamics and variability of turbidite systems; and by chemical, biological and physical studies of fluid flow, gas hydrates and deep-water seeps and biota.
Phase II is envisioned to extend the work off Europe to other margin systems that are also of interest to European scientists. Particular target areas include the conjugate margins of the North Atlantic (e.g., Labrador, Greenland and west Iberia), the young margins of the Gulf of Aden and, the highly segmented margins of the South Atlantic (Brazil, Gabon and the Congo).

9.4 InterMARGINS

Over the past few years, continental margins research has become a major focus of the international geoscience community. New national research programs have been initiated in many countries, for example in France, UK, USA, and Japan. To foster a greater degree of international coordination of margins research activities, to focus sufficient resources on some common, large interdisciplinary investigations, and to help leverage funding for multi- and inter-disciplinary research projects, a new international geoscience initiative dedicated to continental margins research was recently formed. The result is a steering committee, termed InterMARGINS, that is an international and interdisciplinary group of researchers concerned with coordinating continental MARGINS research between the various national MARGINS programs. It is designed to encourage scientific and logistical coordination, with particular focus on problems that cannot be addressed as efficiently by nations or national institutions acting along or in limited partnerships.

The scientific planning and direction of InterMARGINS is the responsibility of its steering committee and various InterMARGINS working groups. Each country with a paid subscription to InterMARGINS is represented by a voting member on the steering committee. The role of the steering committee will be to: 1) define and at times update the program plan, 2) propose and oversee specific InterMARGINS projects, 3) consider and prioritize proposals for new program elements, workshops and other appropriate activities, 4) liaise with the leaders of national MARGINS programs, 5) determine the membership in InterMARGINS and its working groups and committees, 6) approve the InterMARGINS budget and oversee the operation of the InterMARGINS office, and 7) select the InterMARGINS chairman and the host country for the InterMARGINS office. The chair of the office is presently Dr. Robert Whitmarsh of the Southampton Oceanography Center (SOC).

The most important role of InterMARGINS is to foster and enhance communication between national MARGINS-related research programs. It will develop and maintain metadatabases of ongoing national and multinational projects and research activities, initiate and carry out workshops, and disseminate information to members through newsletters and other forms of communication.

10. “Rupturing Continental Lithosphere” implementation plan

Study of the RCL initiative requires an integrated effort over a decadal timescale. The emphasis of research activities will evolve over these years, progressing from early integrated data gathering and characterization, experimentation and model formulation to more logistically complex field studies and integrated laboratory and numerical experiments. Later activities will stress synthesis and formulation of improved models for lithospheric deformation. The following timeline is illustrative, the details of actual activities and the exact sequence of events will be dictated by the proposals written and funded and by logistical and programmatic considerations.

Planning workshops for the Gulf of California/Salton Trough and northern Red Sea/Gulf of Suez focus sites were held in Puerto Vallarta at the end of October, 2000, and Sharm el-Sheikh during March, 2001, respectively. It is envisaged that the initial phases of work be concentrated in the Gulf of California and the Gulf of Suez portions of the focus sites with the geophysical characterization of each focus site at a scale and resolution necessary to generate an adequate foundation for future geological, geodetic and geochemical studies.

The second phase of reconnaissance work would represent geophysical characterization of the Salton Trough and the central Red Sea regions of each focus site. In all cases, this characterization should include the production of detailed topographic datasets obtained by merging onshore DEM information with offshore seafloor swath mapping, regional and high-resolution multichannel reflection seismic data, measuring surface heatflow variations across the focus sites, using lithospheric tomography to map possible lower crustal and lithospheric mantle flow regimes, regional refraction surveys to determine crustal and mantle velocities and Moho geometry, and the construction of regional but high resolution magnetic and gravity grids of onshore and offshore regions to define sub-surface structural trends, depocenter location and subbasin segmentation, and as a critical constraint on the spatial variations in the flexural strength of the extended lithosphere.

This framework will be crucial for guiding subsequent large-scale geological and tectonic syntheses, onshore geological mapping of syn-rift stratigraphy, border fault development and segmentation, continued radiometric dating of rift-related volcanics, and ultimately, the identification of crucial sites for testing hypotheses using dredging and IODP drilling.

Other critical activities in the early stages of the RCL program include developing databases for rapid dissemination of information, establishing local seismic arrays to define the depth-distribution and focal mechanisms of crustal deformation, and establishing geodetic monitoring stations. Theoretical institutes and workshops will help to inform, educate and reflect on the evolving RCL program, facilitating the construction of both a well-informed community knowledgeable of RCL objectives and feedback to the steering committee concerning the need for possible midcourse changes of the program. At the minimum, a mid-term review of the entire RCL program and science plan will be held as a public workshop. During the final years of the program, shallow-water and on-riser drilling and subsequent data analysis will ultimately test predictions about the rheological zonation of the crust and mantle, the partitioning of strain in the deforming lithosphere, and the local and far-field effect of stresses in driving lithospheric extension.

Numerical modeling and laboratory experiments will be critical throughout the program to interpret and synthesize the causes and consequences associated with the spatial and temporal variations of strain through the lithospher. These models will necessarily be constrained the results from the various observational datasets, both regional and local. This marriage of theory with observation will be a fundamental aspect of RCL numerical experiments.

A comprehensive study of the RCL initiative is estimated to cost $20-25 million, exclusive of ship time and drilling costs and facility-related projects associated with geodetic, land, and ocean bottom seismometers. The probability of international cooperation with collaborating organizations and synergism and input from EuroMARGINS and InterMARGINS should help reduce data acquisition and field logistic costs. It is believed that the synergism of nested and focussed multidisciplinary studies should yield a greater benefit/dollar investment that the usual single approach. Having said this though, it is expected that RCL-related proposals submitted to NSF core programs will provide invaluable complementary information on either the themes of RCL or insights from studying other (i.e. non-focus site) margin systems.

11. How will we communicate results and opportunities for cooperation?

A RCL web site is maintained as part of the duties of the MARGINS Office with the prime purpose of distributing MARGINS-related results and data sets significantly prior to publication. The RCL web site will provide the following: 1) information concerning upcoming field expeditions and experiments, 2) access to databases and recently acquired data, and 3) a news bulletin board to foster communication across the different disciplines in the RCL community in addition to listing the abstracts of published work supported by MARGINS funds.

By making information concerning upcoming cruises and field experiments available in a timely fashion, other researchers can capitalize on these opportunities and secure funds to participate in these projects or design piggy-back experiments. In addition, acquired RCL data and/or pathways to access data will be available on the web soon after acquisition. The actual timeframes for data availability will vary for different data types. Existing data from RCL focus sites will be compiled, catalogued and entered on the web database. Rapid dissemination of data and new ideas will help focus the community which in turn should lead to a more interdisciplinary approach towards studying RCL. Data availability, together with the news bulletin board, will improve communication between observationalists, experimentalists and theoreticians.

Theoretical institutes will similarly allow interdisciplinary interaction. Enhanced communication will allow rapid determination of the critical observations and experiments necessary for constraining models for RCL. In this way, first-order model predictions can be immediately tested. Such an iterative approach between modeling and data analysis is the necessary first step towards developing realistic quantitative models of RCL.
In addition to the web site, international meetings and publications will promptly communicate the results of RCL. Workshops on the main scientific themes will also be an important component to focusing efforts within the broad community. These workshops will bring together the diverse community of researchers needed to cross-fertilize ideas and develop multi- and inter-disciplinary approaches to make progress on the RCL scientific objectives. Further, topic-specific workshops will be augmented by AGU and GSA town meetings and special sessions, by Theoretical and Experimental Institutes and associated publications, and by mid-program planning workshops.

Finally, an important component of RCL science execution and communication will be in the training of US and foreign students and developing exchange programs with foreign collaborators. Funds for this research and training will be earmarked in the overall MARGINS budget and will be distributed via successful student and foreign proposals to NSF-MARGINS.

12. How will “Rupturing Continental Lithosphere” be evaluated?

We anticipate that the RCL initiative will progress through a series of peer-reviewed and competitive proposals. This funding mechanism, that of funding the best science as defined by peer review and NSF Panel deliberations, provides continual evaluation of the initiative. Further, the initiative will be compared against the milestones set out for all MARGINS initiatives by the MARGINS Steering Committee in consultation with NSF-MARGINS Program Managers. The milestones will be supplemented by progress reports distributed by the MARGINS Office. These reviews will summarize the research of MARGINS-funded research and some evaluation of the effectiveness of interdisciplinary and international activities. Proposal pressure, MARGINS-related publications and community attendance at RCL workshops and theoretical institutes will all be effective measures of the progress and value of RCL research. A mid-term review in the form of a public workshop will allow the progress of RCL to be assessed in addition to allowing the community to reflect on the progress and possibly redirect/redefine the future direction of the initiative.