To better understand the implications of ridge axis morphology for the thermal and mechanical structure of accreting plate boundaries, we examine mechanisms for the creation of stress-supported axial topography and its dependence on parameters such as spreading rate, asthenosphere viscosity, and plate thickness near the ridge axis. Three basic mechanisms are considered: (1) asthenospheric upwelling in a narrow, steady-state conduit beneath a ridge axis; (2) mantle flow stresses due to plate spreading, extending ridge axis lithosphere, or melt segregation and associated mantle compaction; and (3) topography supported by moments due to horizontal extensional stresses in a thickening lithosphere. If mid-ocean ridge topography is due to asthenosphere rising in a narrow conduit at the ridge axis, then the conduit structure cannot be thermally controlled, as the associated thermal structure has relatively broad, flat isotherms instead of a conduitlike shape. Mantle flow associated with plate spreading alone induced no stress-supported topography. Thus the intuitive notion that diverging flow at the ridge axis due solely to plate spreading produces the axial valley is incorrect. Vertical mantle flow at the base of a stretching, accreting lithosphere and mantle flow associated with melt segregation can, however, lead to stress-supported axial topography. However, for these mantle flow related processes to produce significant amounts of axial relief, the mantle viscosity beneath the ridge axis must be on the order of 1020 Pa s(1021 P(poise)), is significantly higher than generally accepted values of 1018 Pa s (1019 P).

Horizontal extensional stresses in a strong, brittle lithosphere that thickens with distance from the ridge axis can produce axial topography of the form and amplitude observed on slow spreading ridges. A continuum idealization of deformation due to faulting is used to examine this mechanism. This model predicts the axial valley width is controlled by the plate thickness near the ridge axis. An 8-km-thick plate near the ridge axis would produce the 30-km-wide axial valley observed at slow spreading ridges. An increased lithosphere thickness of only several kilometers over the half-width of the axial valley can produce kilometer-scale relief. To examine whether a plate 8--10 km thick at the ridge axis that thickens by several kilometers within 15 km of the axis is plausible at a slow spreading spreading ridge, thermal models that include both the heat of magnetic crustal accretion on the ridge axis and upwelling mantle flow beneath the axis are formulated. Hydrothermal cooling that increases the effective conductivity by about a factor of 10 is required for an 8-km-thick plate to exist at a slow spreading ridge axis.

However, even with this enhanced hydrothermal cooling, the plate thickness at a fast spreading ridge axis and the resulting horizontal force in the plate are too small to result in appreciable topography. While mid-ocean ridge axial topography is strongly influenced by spreading rate, other factors also shape axial structure. For example, the anomalously shallow slow spreading Reykjanes Ridge has an axial structure more like that of a fast spreading ridge, and the anomalously deep intermediate spreading Australian-Antarctic Discordance has a median valley axial structure typical of a slower spreading ridge. In addition, within a given spreading segment there is often a systematic deepening of the ridge axis toward transform offsets. These non-spreading-rate-dependent axial topographic variations may be produced by variations in magmatic heat input associated with crustal emplacement which shapes the thermal structure and lithosphere thickness near the ridge axis. ¿ American Geophysical Union 1987