Rifting and seafloor spreading in backarcs: The Bransfield and North Fiji Basis (NW Antarctica and SW Pacific)
Abstract
This Thesis deals with the study of the rifting and seafloor spreading processes in backarc basins using swath-bathymetry and geophysical data. The geophysical methods used are mainly magnetics and gravity, although other methods, such as seismic reflection and submersible data are also presented in some chapters. Three areas were selected for this study: The Central and Eastern Bransfield Basins, northwest Antarctic Peninsula, and the Central Spreading Ridge and the South Pandora-Tripartite Ridges in the North Fiji Basin, southwest Pacific.
The Bransfield Basin is a narrow and elongated active backarc basin located between the Antarctic Peninsula and the South Shetland Islands, at the southwestem edge of the Scotia Arc (Fig. 1 a). The Bransfield Basin is composed of three small basins, Western, Central and Eastern, separated respectively by Deception and Bndgeman Islands. The last two were surveyed by the GEBRA 93 cruise during which full swath-bathymetric coverage, single-channel seismic reflection and magnetic profiles were acquired.
The Central Bransfield Basin (Fig. la) is 60 km wide, 230 km long and 1950 m deep, and the structures mainly trend N55-60. The basin morphology is dominated by six large seamounts (labelled A to F) that crop out from the sedimented seafloor of the Central Bransfield Basin and align with the basin axis. The seamounts present circular, semi-circular and elongated morphologies. Moreover, the Eastern Bransfield Basin (Fig. la) is 42 km wide, 150 km long, deeper than 2700 m and trends N40-45. The basin is charactensed by four deep en échelon troughs showing a lozenge shape, and small, scattered volcanic cones mainly located in the southwestern half basin. A total of 119 submarine volcanoes are observed in these two basins, with a predominance of higher edifices (over 150 m high) in the Central Basin. Magnetic anomalies are difficult to identify in the Bransfield Basin, although a positive alignrnent well correlated with the submarine volcanic edifices of the Central Bransfield Basin was recognized and named the Bransfield Rift Anomaly. When this anomaly is tentatively interpreted as Anomaly 1, the maximum age of spreading in the Central Bransfield Basin would be 0.71 Ma and the resulting maximum full rate 0.83mm/year. The free-air gravity anomalies are well correlated with the bathymetric maps.
The North Fiji Basin is a mature backarc basin located between two active subduction zones of opposite polarity: the New Hebrides and the Tonga-Kermadec trenches (Fig.lb). Severa1 extensional features and spreading centres have been identified within the North Fiji Basin. Two of these features are studied in detail here: the Central Spreading Ridge and the South Pandora-Tripartite Ridges.
The Central Spreading Ridge (Fig. lb) is the most widely explored and best known of al1 the spreading centres identified in the basin, and has been intensively explored during the cruises of the French-Japanese STARMER project (1987-1991). Six cruises were undertaken in the area during this project: 4 surface cruises (swath-bathyrnetry, geophysics and sampling) and 2 diving cruises. The Central Spreading Ridge is more than 800 km long and 50-60 km wide, and is segmented into three first order segments labelled, from north to south, N160, N15 and N-S according to their orientation. The N160 segment is 210 km long and is composed of three second-order segments (CSR1 to CSR3), which are a succession of long en échelon grabens, similar to that of the Mid-Atlantic Ridge. The N15 segment is 165 km long and comprises two segments, CSR4 and CSR5. The axial morphology is characterized by a double ridge, split by an axial graben. The N-S segrnent is 255 km long and divided into three second-order segments (CSR6 to CSR8). The segment morphology shows a central flat and rectangular high, like in the East Pacific Rise. Magnetic anomalies up to Anomaly 2A (3.5 Ma) are recognized along the N-S segment, whereas only Anomalies J and 1 (0.97 and 0.7 Ma, respectively) are clearly identified along the other two segments. The calculated spreading rate is intermediate, decreasing northwards from 80 to 50 mm/yr. In addition, there is a change in the axial gravity structure along the Central Spreading Ridge. The mantle Bouguer anomalies obtained on the northern part of the Central Spreading Ridge (N1 60fN 15 segments) show "bull's eye" structures interpreted as the result of mantle upwelling at the middle of the segments. In contrast, the mantle Bouguer anomalies of the southern part of the ridge (N-S segment) are more homogeneous and consistent with the observed smooth topography.
The South Pandora-Tripartite Ridges (Fig. lb) are located in the northern part of the North Fiji Basin, one of the less explored areas of the basin. The area was surveyed by the NOFI cruise during which swath-bathymetry, geophysical and geological data were acquired. The South Pandora-Tripadite system extends over more than 800 km, and three first-order segments are distinguished according to their orientation: N75 and E-W (South Pandora Ridge) and N1 10 (Tripartite Ridge). The N75 segment is 170 km long and is composed of two second-order segments, SPR4 and SPR3. Its axial morphology is dominated by an axial graben disturbed only by an abrupt central seamount in the middle of segment SPR3. The E-W segment is 300 km long and composed of three segments, SPR2, SPRl and SPRO, from west to east respectively. This segment also shows a contrasting longitudinal morphology. The N1 10 segment is 280 km long and composed by TR3, TR2 and TR1 second-order segments. Bathymetric maps show axial seamounts and deep troughs alternating along-axis. Preliminary interpretations of magnetic anomalies are presented, and Anomalies 1 to 3A (7 Ma) are identified along the South Pandora Ridge. The calculated spreading rate is ultra-slow at 16 dyr. In contrast, only Anomaly 1 is clearly identified along the Tripartite Ridge, with a full preading rate decreasing from 8.5 to O rnmíyr towards the southeast. The gravity structure also shows "bull's eye" lows associated with the contrasting volcanic highs.
Severa1 points are discussed concerning backarc evolution, the comparison between backarc and midocean ridge spreading, the role of seamount volcanism, differences in thermal regimes along the Central Spreading Ridge, and models of backarc accretion. Themain conclusions are:
1) The three areas studied have been classified in terms of backarc evolutionary stages, from incipient and prespreading rifting stages (Central and Eastern Bransfield Basins, respectively) to well organized seafloor spreading, and from young (Tripartite Ridge) to mature (Central Spreading Ridge and South Pandora Ridge).
2) The present-day opening seems to be related to the rollback of the subduction hinge in the Bransfield Basin. In the North Fiji Basin the opening seems to be linked to a regional thermal anomaly.
3) The initial rifting of the arc may pre-determine the future segmentation of the basin. Even if the accretionary processes are similar in backarc and mid-ocean ridge settings, differences appear concerning ridge segmentation and axial discontinuities. First-order backarc segmentation is short-lived and about half the length of that in mid-ocean ridges. A fundamental dif difference between backarc and mid-ocean ridges is the lack of large fracture zones and transform faults separating the backarc segments.
4) Large seamount volcanism may play a fundamental role in backarc axial construction, and may be a common characteristic of slow and ultra-slow mature spreading ridges and incipient spreading centres, as observed along the South Pandora-Tripartite Ridges and the Central Bransfield Basin, respectively.
5) The variability of the axial morphology and gravity structure observed along the Central Spreading Ridge is explained in terms of differences in thermal regime. The limits between "cold" and "hot" segments are propagating rifts, which may be interpreted as thermal boundaries.
6) Two end-member models of backarc crustal accretion and mantle upwelling are presented: focused-type and continuous-type accretion. The focused-type would show an extremely contrasted morphology and deep structure along the segments, with punctiform upwe-Ilings. The continuous-type accretion would be homogeneous and uninterrupted along the ridge, with a persistent magma chamber. Interna1 (ridge evolution, spreading rate, thermal regime) and externa1 factors (nature of the surrounding lithosphere, proximity of mantle plurnes and subduction zones) may control the different types of accretion.
