Systematic changes in seafloor depth, crustal structure, and crustal geochemistry can occur within 30 km of a fracture zone. The seafloor gradually deepens roughly 1 km within 30 km from the fracture zone, the crust may be thinner than crust created far from a fracture zone, and systematic compositional differences are observed between basalts erupted near and far from fracture zones. We call these effects the transform fault effect (TFE). To investigate the physical Processes responsible for the TFE, a general numerical method is developed to solve for the three-dimensional flow and thermal structure beneath a mid-ocean spreading center. This model is applied to study an idealized spreading center consisting of a 100 km transform fault offsetting two ridge segments spreading at rates of 1, 2, and 4 cm/yr. Using an adiabatic melt relation and our flow and temperature calculations, we find the distribution of melt production beneath the spreading center. Finally, a porous flow model of melt migration within a spreading center is developed to assess the possible effects of melt migration on oceanic crustal structure. We find that the expected topographic effects caused by mantle density variations associated with cooling near a transform offset are far smaller than the observed 1 km seafloor deepening within 30 km of a fracture zone. While significant compositional upper mantle heterogeneity due to lower degrees of partial melting will be associated with fracture zones, isostatic compensation of ridge parallel upper mantle heterogeneity also does not seem a likely cause for observed seafloor deepening toward a fracture zone. Crustal thickness variations is a good candidate to explain this seafloor deepening.

Several kimometer of crustal thinning can plausibly occur within 30 km of a fracture zone. This thinning is due to both perturbations in melt migration at a transform offset with melt preferentially migrating toward the center of a spreading segment and perturbations in melt production near a transform offset caused by lower upwelling rates near the transform. The major influence of a transform offset on melting beneath a ridge-transform spreading center is due to the muting effect of a transform offset on upwelling beneath the ridge-transform intersection; lower rates of upwelling lead to lower amounts of melt production within a broad region near the transform. Our results (weakly) support Bender, Langmuir, and Hansen's petrological investigations of along-axis variability in the depth of melting. They found a greater average depth of melting for basalts erupted toward the center of a spreading segment versus basalts erupted near a transform fault. Our results suggest that the actual depth of the beginning of melting along a spreading center is not strongly affected by a transform offset but that melt production occurs over a larger depth interval toward the center of a spreading segment. Thus melt production variations alone tend to produce a trend opposite to that found by Bender, Langmuir, and Hansen. However, variations in melt migration can produce the trend found by Bender, Langmuir, and Hansen; melt created at greater depths beneath a transform will migrate further toward the center of a spreading segment than melt created at shallow depths. Present calculations are inconclusive in showing whether the effects of melt production or melt migration dominate at a ridge-transform intersection. ¿ American Geophysical Union 1988