Many areas along the Mid-Atlantic Ridge (MAR) have been surveyed in considerable detail. Surprisingly, the part of the Reykjanes Ridge connecting to Iceland is not one of these areas, although this is arguably the most important part of the ridge system to survey to understand plume-ridge interactions. The position of Iceland over the Mid Atlantic Ridge (MAR) provides an ideal setting in which to investigate mid-ocean ridge processes and the effects of hotspots on these processes (e.g. RIDGE 2000 Integrated Study Site Proposal for Hotspot-Influenced Oceanic Spreading Centers: Iceland and the Reykjanes and Kolbeinsey Ridges). Iceland is one of only two places on Earth where an oceanic spreading center rises above sea level, which allows nearby work on the submarine ridges to be placed in the extensive geological and geophysical context established for subaerial Iceland. These studies have established the basic pattern of present-day kinematics, geochronology of plate boundary shifts, and geochemical characteristics of the plume, and demonstrate strong plume interaction with the MAR at the Reykjanes Ridge.
The Reykjanes Ridge (RR) separates the North American and Eurasian plates in the north Atlantic south of Iceland (Fig. 1). It is the longest oblique spreading ridge in the world, extending about 900 km from Iceland to the Bight transform fault near 56.5°N. The Bight transform is a 15 km left-stepping displacement of the ridge axis along 092º, with an associated fracture zone extending into at least 36 Ma crust (Müller and Roest, 1992), that marks a change in ridge orientation from near orthogonal to the south to highly oblique in the north. On average, the RR trends ~036° with a spreading direction of ~099°, representing an average spreading obliquity of 27° (Vogt et al., 1971; Fleischer, 1974; Keeton et al., 1997). However, there is an increase in obliquity towards Iceland, reaching a maximum on the Reykjanes Peninsula, where it is ~065° (Johnson and Jakobsson, 1985). The average spreading rate along this slow spreading ridge is about 20 mm/yr (Talwani et al., 1971, DeMets et al., 1994). The RR is unusual in that it lacks any transform faults, and also in terms of its general physiology. South of ~59°N the RR axis forms a prominent axial valley ~40 km wide and ~1 km deep, typical for slow spreading ridges not influenced by hotspots. Closer to Iceland the ridge axis shallows and the axial valley is replaced by a robust axial high more characteristic of a fast-spreading ridge (Laughton et al., 1979; Murton and Parson, 1993; Parson et al., 1993; Searle et al., 1998; Fig. 2). In general this is attributed to effects of a mantle plume currently situated under Iceland (e.g. Morgan, 1971; Vogt, 1971, 1974; Schilling, 1973, 1986; White et al. 1995; White, 1997; Wolfe et al., 1997; Allen et al., 1999; Allen, 2001).
Distinctive diachronous V-shaped ridges are observed in the gravity (Fig. 1) and bathymetry (Fig. 2) data, progressively terminating transform faults in what had been a more typical orthogonal ridge/transform system (Talwani et al., 1971; Vogt, 1971). These V-ridges have thicker crust than normal (White et al., 1995; Weir et al., 2001; Smallwood and White, 2002), consistent with formation at higher mantle temperatures, and there is a strong plume geochemical signature along much of the RR (e.g. Schilling, 1973, 1975; White et al., 1976; Schilling et al., 1982). That these V-ridges are associated with the Iceland hotspot and possible plume seems inescapable, although exactly what the relationship is remains a subject of debate. They have usually been interpreted to result from “pulses” of the plume and subsequent subaxial flow of magma (e.g., Vogt, 1971; Vogt and Johnson, 1975; White et al., 1995; Ito, 2001), although there is an alternative hypothesis involving ridge jumps (Hardarson et al., 1997) that we propose to test.
The RR rises above sea level at the southwestern tip of Iceland at the Reykjanes Peninsula (63.8°N). The island of Iceland straddles the MAR and is the largest area of sub-aerially exposed mid-ocean ridge on Earth. Over Iceland the MAR is spreading at ~19 km/m.y. along an average direction of ~110°. The unusually high rate of volcanism that generates the Iceland bathymetric high can be related to the combined contributions from the Iceland hotspot and the mid-ocean ridge (Morgan, 1971; Schilling, 1973). The geometry of the present day MAR plate boundary through Iceland is complicated (Fig. 3), with spreading occurring on 3 principal rift zones, the Western (Reykjanes-Langjökull) Volcanic Zone (WVZ), the Eastern Volcanic Zone (EVZ), and the Northern Volcanic Zone (NVZ), with two principal transform zones, the South Iceland Seismic Zone (SISZ) and the Tjörnes Fracture Zone (TFZ) (Ward, 1971; Saemundsson, 1974, 1979; Jacobsen, 1979; Björnsson et al., 1979; Jóhannesson, 1980; Gudmundsson, 1995; 2000; Hardarson et al., 1997; Thordarson and Hoskuldsson, 2002).
It has been established that the position of the plate boundary has shifted repeatedly to the east during the 16 Ma history of present day Iceland (e.g. Ward, 1971; Burke et al., 1973; Saemundsson, 1979; Johannesson, 1980; Aronson and Saemundsson, 1975; Helgason, 1984; Vink, 1984; Thordarson and Hoskuldsson, 2002). The rift zones evidently adjust to the westward movement of the MAR away from the center of the Iceland plume, now thought to lie below the northern part of the Vatnajökull glacier (Johannesson, 1980; Ryan, 1990; Gudmundsson, 1995; Wolfe et al., 1997; Shen et al., 1998; Allen, 2001), by episodically jumping east to stay near the hotspot. This explains the offset of the EVZ/NVZ from the RR/Kolbeinsey ridge system, and the growth of the SISZ and TFZ (Ward, 1971; Fig. 3). In northern Iceland, there was a major eastward ridge jump at ~15-16 Ma from Vestfirdir, in the extreme NW of Iceland, to the Skagi Peninsula, and another beginning ~7 Ma from Skagi to the NVZ (Saemundsson, 1974, 1979; Helgason, 1985; Hardarson et al., 1997). In southern Iceland there was a rift relocation from the Snaefellsnes Peninsula (Fig. 3) to the Reykjanes Peninsula ~6-9 Ma (Hardarson et al., 1997; Kristjansson and Jonsson, 1998). The most recent plate boundary shift, the ridge jump thought to be in progress from the WVZ to the EVZ beginning some 2-4 Ma ago (Johannesson, 1980; Gudmundsson, 1995), formed the southern EVZ that extends from the Vatnajökull glacier in the north to Surtsey and the Vestmann Islands off the south coast of Iceland, the latter possibly at the tip of a southward propagating rift (Meyer et al., 1985; Saemundsson, 1986; Mattson and Hoskuldsson, 2003). Instead of occurring instantaneously, these shifts occurred gradually over a period of a few m.y. while spreading was distributed over both the old and new rift zones.
The hypotheses we propose to test are that there were rift relocations (ridge jumps) along the RR that were associated with the known rift relocations on Iceland, and that these jumps are associated with the V-shaped ridges extending south from Iceland.
The data necessary to test these hypotheses are magnetic anomalies, gravity, and multibeam bathymetry across the part of the RR closest to Iceland (N of 62°N), collected along the flowline of relative plate motion to minimize complexities that ridge jumps produce, and maximize the signal/noise ratio of these jumps to facilitate our modelling effort.
The test is straightforward. We will model each magnetic profile using our Magbath program (Fig. 10) as well as modelling the contoured magnetics using our PRMap program (e.g. Fig. 8) to determine whether the observed asymmetry in the V-shaped ridges results from small ridge jumps or from asymmetric spreading with no resolvable jumps. If the V-shaped ridges are associated with ridge jumps, and if these jumps occur in patterns characteristic of propagating rifts (Fig. 11), they will lie along pseudofaults or failed rift sequences forming part of propagating rift wakes that will be revealed by our data and modelling. If the asymmetry does not result from jumps, there will be no failed rifts or pseudofaults.
If there were jumps, we will determine how the pattern relates to the known pattern on land, to produce a seamless history of the North America-Eurasia plate boundary geometry on and near Iceland during the past 18-20 Ma. Our results would almost certainly add to the understanding of the rift relocations on Iceland, where the detailed ridge jump history is buried by lavas from the eventually successful plate boundary.
If there were no jumps along the RR, we will determine what it means for the regional tectonic evolution to have had ridge jumps on Iceland but not the RR, and how this has influenced the evolution of Iceland, e.g. the predictable growth and increasing obliquity of the Reykjanes Peninsula and South Iceland Seismic Zone. This would constrain geodynamic models of processes such as sublithospheric magma flow.
A comparison between the “flow down a pipe” plume pulse model (e.g. Vogt, 1971), and the propagating rift model (Fig. 9), shows one essential difference that will allow us to test between these hypotheses. If the seafloor spreading process itself is symmetric, as commonly assumed because of the extreme dependence of lithospheric strength on temperature (Morgan, 1968), the simple pipe flow model (Fig. 9A) predicts symmetric accretion of lithosphere to the 2 plates. In contrast, the propagating rift model (Fig. 9B) requires the transfer of lithosphere between plates, and thus predicts systematic asymmetric accretion, such as that seen in the Talwani et al. (1971) profiles (Fig. 5).
We propose a marine geophysical survey to determine the 0-20 Ma seafloor spreading history of the part of the RR nearest Iceland. The critical areas of seafloor flanking the RR have not been previously surveyed systematically by ship, making it currently impossible to determine if there is a pattern of oceanic ridge jumps that should relate to those observed on Iceland. It is important to do this to test between competing hypotheses for the major V-shaped patterns of bathymetry and gravity, and to extend the documented effects of plume-ridge interaction to more peripheral locations relative to the plume center. Although various “pulsing plume” hypotheses to explain this pattern have long been favored, recently an alternative “rift relocation” hypothesis has been proposed. The Hardarson et al. (1997) hypothesis argues that the RR is normally very elevated, with the V-shaped ridges actually being defined by the valleys between them which result from magmatic deficiency during times that ridge jumps prevent some of the plume magma from reaching the RR. It thus predicts some sort of strong correlation between the ridge jump pattern and the V-ridge pattern, although exactly what the correlation should be in detail is uncertain. What is certain is that if there is a pattern of offshore jumps that correlates with the onshore pattern, it will be possible to use the seafloor data to understand the detailed evolution of Iceland much better, because ridge jumps on Iceland are associated with voluminous volcanism that overwrites and deeply buries the preexisting crust, whereas the seafloor record of jumps is written and preserved on a much finer scale. If correlations are not found between the seafloor and Icelandic patterns of jumps or between the jumps and the V-shaped ridges, this would argue that some sort of plume pulse mechanism is responsible for the V-ridges rather than the Hardarsson et al. hypothesis. Of course hybrid hypotheses involving plume pulses causing ridge jumps are also possible and will be constrained by our data. Our goal is a seamless history of the plate boundary geometry both at sea and on land. We regard this as an essential step toward the full understanding of Iceland and the geodynamic influence of the hotspot or mantle plume on the mid-ocean ridge system.