Geophys. J. Int. (1994) 116, 64-84 A global analysis of mid-ocean ridge axial topography Christopher Small Scripps Institution of Oceanography, UCSD, La Jolla, CA 92093-0208, USA Accepted 1993 June 15. Received 1993 June 10; in original form 1992 August 20 INTRODUCTION SUMMARY Current views of mid-ocean ridges are strongly influenced by extensive mapping of the Mid-Atlantic Ridge and East Pacific Rise. The global picture of the mid-ocean ridge system, particularly in the sparsely surveyed Southern Oceans, is still based primarily on underway bathymetry profiles collected over the past 40 years. This study presents a quantitative analysis of global mid-ocean ridge morphology based on 156 of these underway bathymetric profiles, thereby allowing commonly recognized features such as axial valleys and axial ridges to be compared on a global basis. An Empirical Orthogonal Function (EOF) analysis is used to separate deterministic and stochastic components of axial morphology and to quantify the dependence of each on parameters such as spreading rate and axial depth. It is found that approximately 50 per cent of the variance in axial morphology may be described as a linear combination of five independent symmetric and anti-symmetric modes; the remainder is considered stochastic. Maximum axial valley relief decreases with spreading rate for rates less than 80 mm yr- 1 while axial ridge relief remains relatively constant for all rates greater than 50 mm yr- 1 • The stochastic component ofAhe axial morphology, referred to as bathymetric roughness, also decreases with spreading rate for rates less than 80 mm yr- 1 and remains relatively constant at higher rates. Although both axial valley relief and bathymetric roughness near the ridge axis show a similar spreading rate dependence, they are weakly correlated at slow spreading centres. The distinct differences in morphologic variability of fast and slow spreading ridges may result from the episodicity of magmatic heat input which controls the lithospheric rheology at slow spreading ridges. These observations support the notion of a critical threshold separating two dynamically distinct modes of lithospheric accretion on mid-ocean ridges. Key words: Mid-Atlantic Ridge, oceanic lithosphere, sea-floor spreading, topography. It is well known that mid-ocean ridge axial morphology varies with spreading rate. The traditional view of the ridge system states that slow spreading ridges are characterized by a 1-2 km deep axial valley bounded by rift valley mountains and rugged flanking topography, while fast spreading ridges are characterized by a more continuous axial ridge, less than 1 km in height and bounded by smoother flanking topography. Numerous multibeam surveys of the northern Mid-Atlantic Ridge and East Pacific Rise have modified the traditional view of the ridge system. It is now known that both fast and slow spreading ridges are considerably more segmented than was previously believed. In particular, the distinction of smaller scale migrating offsets from the larger scale, more stable transform offsets has been recognized. It has also been shown that axial depth and relief often vary in a consistent manner with distance from different types of offset on both fast and slow spreading ridges. Although systematic multibeam surveys provide a more complete view of the ridge axis, the majority of the ridge system has been surveyed only with single wide beam echosounders. As mentioned above, the majority of our knowledge of ridge axis morphology is based on multibeam surveys of the northern Mid-Atlantic Ridge and East Pacific Rise. While these ridges provide convenient access to the fast and slow spreading end members of the system, they comprise less than half of the global mid-ocean ridge system. Current knowledge of the ridges in the Southern Oceans is still based almost entirely on expeditions conducted prior to the mid 1970s (Smith 1992). Because very few multibeam surveys have been conducted on the 64 by guest on May 31, 2016 http://gji.oxfordjournals.org/ Downloaded from