Auxiliary material for Paper 2010GL045165 Lower crustal variability and the crust/mantle transition at the Atlantis Massif oceanic core complex Donna K. Blackman Scripps Institution of Oceanography, University of California, San Diego, La Jolla, California, USA John A. Collins Woods Hole Oceanographic Institution, Woods Hole, Massachusetts, USA Blackman, D. K., and J. A. Collins (2010), Lower crustal variability and the crust/mantle transition at the Atlantis Massif oceanic core complex, Geophys. Res. Lett., 37, L24303, doi:10.1029/2010GL045165. Introduction The auxiliary material contains details on instrument locations on the seafloor, additional record-sections showing data quality, illustrations of tomographic predictions and results using different starting models, and the tomographic results for Lines 9a and 9b. Analysis methods are discussed in the article text. The data reported here were collected during cruise EW9704 and were submitted to the UTIG Academic Seismic Portal in October 2010 (http://www.ig.utexas.edu/sdc/cruise.php?cruiseIn=ew9704). 1. 2010gl045165-ts01.pdf Table S1. Locations of relocated instruments along Line 8. 2. 2010gl045165-ts02.pdf Table S2. Locations of relocated instruments along Lines 9a and 9b. 3. 2010gl045165-fs01.pdf Figure S1. Details of the OBS refraction experiment at Atlantis Massif. Bathymetric contours are plotted at 100-m interval, heavy lines at 500-m interval. Instrument locations are shown by blue triangles. Line numbers and shooting directions are indicated by arrows. Shot numbers (every 5th) are labeled along each line. 4. 2010gl045165-fs02.pdf Figure S2. Refracted arrivals recorded by ORB1 (a) and OBH26 (b) downward continued to the seafloor. Top panels in a and b are plotted with a reduction velocity of 6.5 km/s. The phase velocities of the refracted arrivals are clearly less than mantle values. The kirchhoff downward continuation accentuates noise, producing the short hyperbolic events that curve away from the main arrival in some places, particularly for OBH26. c) Seafloor topography and position of the instruments along the line. Record-sections above are aligned with the corresponding instrument. 5. 2010gl045165-fs03.pdf Figure S3. Example of tests for Line 8 models with sharp Moho. The green and purple curves show predicted arrival times (a) and raypaths (b) of pre-critical and post-critical PmP reflections from a sharp Moho. The reversed portion of the record-section for the same instrument shown in Figure 2 is shown here in gray. The Pn and Pmp arrivals predicted for a sharp Moho are not seen in these data nor for any other Line 8 instrument. At the edges of the model, sparse data allow a step-wise increase but other models show this is not required by the data. The inversion in b has overall c2~1. Minor cleaning up of the picks shown here (dots on travel-time plots) was done prior to inverting for a final model (Figures 2, preferred, and S5). 6. 2010gl045165-fs04.pdf Figure S4. Tomographic models for Lines 9a and 9b. High velocities occur in the Central Dome and lower velocities are determined in the upper ~1.5 km on the southern dome. Dark-shaded areas have no ray coverage. a) Line 9b. b) Line 9a; location of IODP drill hole is indicated. c) Velocity-depth profiles at two similar locations along Line 9b (gray) and Line 9a (black), which is to the east (Figure 1). Dashed- Southern Ridge (arrow, -3.5 km); solid- Central Dome (arrow, +3.0 km). 7. 2010gl045165-fs05.pdf Figure S5. Effect of starting model on tomographic result. Starting model includes structure in the upper km from Canales et al. (2008) and unaltered mantle velocity is specified to start at 2.5 km subseafloor. The main structure in the inversion is similar to that shown in Figure 2 except where ray coverage is sparse at the edge of the model. Color scale for upper panels is same as for Figure S3 and Figure 2. Differences within the OCC are generally <0.1 km/s, as shown in lower panel.