Pacific Coastal and Marine Science Center

Offshore Earthquakes and Landslides

The Chile Margin Triple Junction:
Modern Analog to Ancient California?

by Steve Lewis (former research geophysicist at the USGS)

Introduction

The region of the Chile Trench along the west coast of South America between about 45° 40' S and 47° S is the site of a collision between the actively-spreading Chile Ridge and the Chile Trench subduction zone. The Chile triple junction region is one of only two presently active examples of a ridge/trench collision, an event that has occurred numerous times around the convergent margins of the Pacific basin, most notably along the California continental margin between about 24 million years ago and the present. Scientific investigations of the active Chile triple junction region can provide important observations that will improve our understanding of these important plate tectonic phenomena, especially insights into past processes that may influence the present geological development and earthquake risk of coastal California.

Several research cruises have been conducted in the region of the Chile margin triple junction in the last decade, including a detailed Seabeam swath bathymetric and seismic reflection survey conducted by the R/V R. D. Conrad, operated by the Lamont-Doherty Earth Observatory of Columbia University in early l988 specifically as a site survey for a later Ocean Drilling Program cruise, and GLORIA side-scan sonar surveys in December, l988. The Ocean Drilling Program expedition was conducted between November 1992 and January 1993.

Regional Tectonics and Geology

Image of South Pacific Tectonic Plates

Figure 1: Tectonic sketch map of the South Pacific Ocean, showing the locations of the major plates and their boundaries. The active spreading ridge between the Antarctic plate and the Nazca plate is being subducted beneath the western edge of the South American plate at 46° South latitude.

The Chile margin triple junction represents the only presently active ridge-trench collision where the overriding plate is composed of continental lithosphere. Hence, the Chile margin triple junction provides the best (only!) active modern example of the geological results of ridge subduction along continental margins, a process that has dramatically affected the geology of Tertiary western North America, among other places. Regional plate tectonic reconstructions for the southern Chile triple junction are well-constrained by marine magnetic anomaly studies, so the detailed relationships between plate motions and continental margin geology can be effectively studied here. These reconstructions show that the Chile Ridge first collided with the Chile Trench about 14 mybp near the latitude of Tiera del Fuego. A long ridge segment was subducted between Tiera del Fuego and the Golfo de Penas between roughly 10 and 14 mybp, another ridge segment was subducted adjacent to the Golfo de Penas (and perhaps partially overlapping with the Taitao Peninsula) about 6 mybp, and a short ridge segment was subducted adjacent to the Taitao Peninsula about 3 my ago.

The relative plate motion vectors change considerably following the passage of the triple junction along the margin. Prior to the ridge collision the Nazca plate was being subducted at a rapid rate (roughly 80 mm/yr for the past 3 my, and as fast as 130 mm/yr during the late Miocene) in a direction slightly north of east. Following the passage of the triple junction, the Antarctic plate is subducted at a much slower rate, roughly 20 mm/yr for the past 15 my, in a direction slightly south of east.

New SEABEAM bathymetric data accurately delineate the present-day geometry and location of the ridge-trench collision. North of about 46° 12' S, the Nazca plate is being subducted beneath the South American plate. South of that latitude the Antarctic plate is being subducted beneath South America. The Nazca/Antarctic plate boundary is comprised of the Chile Ridge spreading center, which intersects the Chile Trench at 46° 12' S, forming a ridge-trench-trench triple junction. The spreading ridge strikes nearly parallel to the trench, resulting in a highly oblique ridge-trench collision, while the fracture zones associated with the Chile Ridge spreading system trend within about 20o of perpendicular to the trench. The triple junction region appears to have formed the southern limit of coseismic rupture during the great l960 Mw = 9.1 Chile earthquake.

thumbnail of plate margins

Figure 2: Location map showing the position of the Chile Margin triple junction along the southern margin of South America, and the regional geography of the triple junction region at 46° S. The active ridge segment between the Darwin and Taitao Fracture Zones is presently being subducted beneath South America. Also shown by boxes are the locations of subsequent bathymetric maps of the triple junction region.

Southern Chile: A seismic analogue to Cascadia?

It has been proposed that the Cascadia margin of offshore Oregon and Washington has the potential for a great earthquake based on similarities to strongly coupled subduction zones, of which the southern Chile margin is a classic example. Contributing to this conclusion are the observations that both margins have sediment filled trenches and both are subducting young crust. However, it is not clear how similar the two margins really are and what parameters are critical for comparing the margins. Thus, in order to realistically compare southern Chile with Cascadia it is necessary to learn a lot more about the southern Chile margin.

THE COLLISION ZONE: 46° S to 47° S

Three Stages in Ridge Subduction

The section of the Chile Ridge between the Darwin and Taitao fracture zones is currently passing beneath the landward trench slope.

The Seabeam bathymetry acquired during Conrad cruise C-2901 provides a more detailed picture of the interaction between the ridge and the trench. On the Seabeam map we can follow the ridge axis from the Darwin fracture zone at 45° 52'S south to 46° 08'S. Along this section of the ridge, the axis is characterized by numerous small volcanoes and by a prominent axial magnetic high. The bathymetry on the seaward side of the axis is dominated by a linear sequence of rift valley walls. Based on these uniform bathymetric features of the young oceanic crust, we infer that spreading is occurring here in a fairly normal manner. In a schematic diagram of the collision zone, we refer to this portion of the ridge and trench slope as the "pre-contact zone."

Thumbnail of Bathy image

Figure 3: Colored shaded relief bathymetric image of the Chile Triple Junction region based on SEABEAM swath data acquired by the R/V Robert D. Conrad, showing the locations of seismic lines 745 and 751. The SEABEAM system is a hull-mounted array of acoustic transducers and associated control and recording computers and other electronics that measures the depth of the water in many discreet points ont he seafloor in a wide swath, widdth roughly 80% of the water depth, to the sides of the vessel. This enables scientists to quickly produce extremely accurate bathymetric maps of the seafloor. In regions where the shape of the seafloor reflects active tectonic or sedimentary processes, this kind of data is very important for understanding those processes. Also shown are Ocean Drilling Program Leg 141 drillsites 859, 860, and 861.

The toe of the landward trench slope adjacent to the "pre-contact zone" is characterized by an escarpment varying between 300 and 500 m high. The strike and position of the toe appears to be controlled by the position of the landward rift valley walls of the subducting oceanic crust. Along the northern part of this zone, the trench toe runs parallel to, and roughly 5 km landward of, the rift axis. Near 45¡58'S the toe steps seaward by about 2 km to a position above the crest of one of the rift valley walls, and then again parallels the ridge axis.

Based on the position of the axial magnetic high, we conclude that near 46° 08'S the ridge axis jumps 3 km to the east. From here south to 46° 12'S the axis of spreading is immediately in front of the toe of the trench. It is not clear from the detailed Seabeam bathymetry whether the numerous small mounds and cones at the base of the toe are all axial volcanoes (some may be debris flows), but volcanic processes are sufficiently active along this section to generate a distinct axial magnetic high. Adjacent to the rift axis the landward trench slope forms a 600 m high escarpment. We refer to this section as the "rift contact zone".

At about 46° 12'S the spreading axis, as marked by the line of small volcanic cones, disappears beneath the landward trench slope. The axial magnetic high simultaneously disappears. From 46° 12'S south to the Taitao fracture zone near 46° 30'S, spreading is occurring beneath the landward trench slope. The most spectacular feature along this section of the trench is a steep, 1200 m high, mid-slope escarpment. We believe that this escarpment results from a massive slump of the trench wall along a seaward-dipping normal fault. At the south end of the collision zone there is a deeply incised canyon, which we refer to as North Canyon, with steep walls over 1000 m high. The region from the disappearance of the ridge axis beneath the trench slope south of the Taitao fracture zone we refer to as the "subducted rift zone".

Structure of the Landward Trench Slope

Line 745 crosses the margin in the pre-contact zone. On this line the axis of the rift valley is about 5 km seaward of the toe of the trench. The inner rift valley wall juts up about 2 km seaward of the toe and the next landwardmost rift valley wall is beneath the toe of the trench. The top of the subducting slab can be followed until it disappears beneath the strong continental basement reflector. A prominent kink in the subducting slab corresponds to the position of another rift valley wall. We tentatively identify the lowermost section of the landward trench slope as being an accretionary complex. However, based on the lack of an accretionary complex further south on Line 750 (discussed in the next paragraph), we note that there may be essentially no young accreted material along Line 745 and the region we identify as the accretionary complex may actually be disrupted and deformed older forearc material. Beneath the mid-and-upper slope region we can resolve at least two prominent, relatively undeformed, reflectors. The shallower of these two reflectors (approx. 0.7 s sub-bottom at CDP 2300) can be traced seaward to at least CDP 1400, and perhaps as far as CDP 1250. This reflector, offset by a series of normal faults near CDP 1800, is the event interpreted to be continental basement on other lines in the region. Overlying this reflector is a higher frequency event that is best imaged between CDP 1350 and CDP 1750, where it lies roughly 1.1 s subbottom to roughly 0.6 s subbottom. In places this event appears to be an unconformity surface within the shelf/slope sediment sequence that overlies basement. We speculate that this reflector unit corresponds to material that was deposited in a relatively stable shelf environment.

Line 745 also display a a prominent BSR reflector roughly 0.2 s beneath the lower portion of the trench slope.

THE POST-COLLISION ZONE: 47° S to 48° S

Margin Adjacent to the Golfo de Penas: 6 MY After Ridge Subduction

The Golfo de Penas segment of the margin, where the ridge collision took place roughly 6 million years ago, has several unique characteristics. Most notable is the presence of a large sedimentary basin within the Golfo de Penas embayment. This basin strikes roughly north-northwest, and contains at least 3 km to 4 km of sediment. The oldest sediment is believed to be Eocene, but the basin has not been drilled. The basin, however, is actively subsiding. It is not clear how the basin responded to the passage of the triple junction, although it almost certainly experienced elevated thermal gradients.

Thumbnail of seismic data

Figure 4: CDP seismic line 745, which crosses the forearc region north of the present-day location of the triple junction. Here the Nazca plate is subducting to the east at about 70 mm/yr. A substantial volume of Pliocene-Pleistocene accreted sediment is present between the trench axis and the seawardmost edge of continental crust can be identified here on the basis of both the seismic interpretation and the drilling results. This section has been pre-stack depth migrated by Dr. Nathan Bangs at the GEOMAR Institute of Kiel University, Germany.

The landward trench slope seaward of the Golfo de Penas has a well-developed accretionary complex (Line 769, see figure for location). This section is ideal for studying the early stages in the development of an accretionary prism and the appearance of an outer arc structural high. If we make the assumption that at the time of the ridge-trench collision the landward trench slope looked something like the landward trench slope further north along the Taitao Peninsula segment, then much of the the accretionary prism, the forearc basin, and the transition zone between them are all features that have developed within the past few million years against a backstop that survived the passage of the triple junction.

CDP Line 769 exhibits most of the structural and morphological features of accretionary subduction z ones. The trench at this latitude contains roughly 2.4 s of flat-lying, undeformed laminated sediment over the subducting oceanic crust. The frontal fold at the toe of the accretionary prism structurally overlies a thrust fault that displays clear listric geometry, and appears to sole out near the top of oceanic crust. This observation implies that most of the trench sediment is presently being accreted to the front of the margin, and that little, if any, sediment is being subducted beneath the accretionary wedge. The seismic image suggests several landward-vergent folds and thrust faults in the middle of the trench slope (CDP 2250 - 2350 and CDP 1850 - 1950), but most structures are seaward vergent. The outer arc high is strongly deformed, but the seismic image resolves several tight folds and small perched sedimentary basins there. A relatively undeformed forearc basin abuts the landward flank of the outer arc high, and contains at least 2.4 s of sediment. A landward-vergent thrust fault system separates the outer arc high from the relatively undeformed forearc basin strata.

Thumbnail of seismic data

Figure 5: CDP seismic line 751, which crosses the forearc region virtually at the location of the triple junction. There is only a very small Pleistocene accretionary wedge at this position along the margin, with continental basement extending almost to the trench. This display is post-stack time migrated.

Thumbnail of bathy perspactive view

Figure 6: Perspective contour map of the Chile triple junction region showing the anticlinal folds of the area south of the triple junction, and the location of Line 769. The seafloor physiography of the margin south of the Taitao Ridge is completely different in style from that north of the Taitao Ridge. This difference is the result of active subduction accretion south of the triple junction and active subduction erosion at and north of the triple junction.

Summary

The Chile Margin triple junction is the best modern example of the subduction of an active spreading ridge at a continental subduction zone, and thus is as close as the modern world offers to what happened along the west coast of North America over the last 20 million years of so. The geologic effects of ridge subduction can be studied relatively easily in Southern Chile because they are occurring as we look, while in California it is necessary to try to look through many million years of subsequent geologic events to try to identify the effects of ridge subduction.

Because ridge subduction represents a huge change in the thermal structure of the continental margin, it has lasting effects on the structure of the crust where it has taken place. These changes might influence such improtant modern phenomena as earthquake seismicity. If we can gain insights in to the geological and geophysical evolution of continental margins by studying active modern examples such as Southern Chile where spreading ridge subduction is taking place, we might be able to better understand the evolution of the San Andreas fault and other regions where similar events have occurred in the past.

Thumbnail of seismic data

Figure 7: CDP seismic line 769, which crosses the forearc region south of the triple junction. Here the Antarctic plate is subducting at about 20 mm/yr. A large accretionary wedge is present at this position along the margin, with a width of at least 30 km. Most of this material has probably been accreted to South America since the passage of the triple junction, less that 2 million years ago. This seismic section has been post-stack time migrated by Dr. Barrie Taylor.

 

U.S. Department of the Interior | U.S. Geological Survey
URL: http://walrus.wr.usgs.gov/research/sopac.html
maintained by PCMSC Web Team
last modified 1998