No abstract

Observed variations in across‐axis topographic relief and faulting style at spreading centers have been challenging to explain. Axial highs are seen at fast‐spreading centers, while valleys occur for slow‐spreading centers. Fault offsets range from tens of meters at fast‐spreading ridges to tens of kilometers at some slow‐spreading ridges. Models that fit the axial relief fail to produce observed fault patterns, while models that fit the fault patterns fail to produce observed variations in axial relief. A recent mechanical analysis (Liu & Buck, 2018, https://doi.org/10.1016/j.epsl.2018.03.045) suggests that including the effect of many discrete diking events can result in a gradual change in axial relief with crustal thicknesses. To compare this mechanical model directly with observations requires us to couple it with a two‐dimensional thermal model. This allows us to estimate the axial lithospheric thickness consistently as a function of the spreading rate and crustal thickness. For thinner axial lithosphere the model predicts an axial high with relief supported by low‐density material beneath the axial lithosphere. For axial lithospheric thickness between approximately one half and approximately three fourths of the crustal thickness, the axial depth decreases with magma supply increase. For thicker axial lithosphere the axial valley relief is controlled by axial brittle lithospheric thickness and near‐axis lithospheric geometry. We compared model predictions to data by compiling observations on axial relief and faulting mode for all spreading centers where seismic crustal thickness has been measured. Good fit to the data is obtained for model parameters giving dike widths in the axial lithosphere close to a meter.