Title: Detection of Subsurface Supergranulation Structure and Flows from MDI High-resolution Data using Time-Distance Techniques. Authors: T.L. Duvall Jr.(NASA/GSFC), A.G. Kosovichev(Stanford), P.H. Scherrer(Stanford), P.N.Milford(Parallel Rules, Inc.) The supergranulation is seen at the surface of the sun in the doppler shift of spectrum lines as an apparent cellular convection pattern with a scale of about 4% of the solar radius. This scale is about 30 times larger than the granulation, seen in white light. Why these distinct scales would be present (and possibly a third intermediate scale mesogranulation) is somewhat of a mystery. Also unknown is the depth structure of the convection. We have used acoustic wave measurements from the MDI experiment on SOHO to address these questions. By crosscorrelating the signal at one location with that on annuli centered on the location, it is possible to measure times for waves to travel over known subsurface ray paths. With some variations on this theme, it is possible to measure horizontal and vertical flows and sound speed variations. Of course, the resulting measurements refer to quantities integrated along these ray paths. An inversion technique based on Fermat's principle has been developed and used to map the flow velocities and sound speed variations as a function of horizontal position and depth. The MDI experiment on SOHO makes doppler shift maps with 1Kx1K points in two choices of image scale, 2 and 0.6 arcsec/pixel. For the present study, we have used the higher resolution mode to observe 8.5 hours of doppler maps sampled once per minute. In order to average enough crosscorrelations to see time-distance effects, the resultant time-distance maps are reduced in resolution by a factor of 10 from the initial data. This still yields about 7 samples across a single supergranulation cell, or 49 over the area of a square cell. Our initial inversions based on the ray theory suggest that the supergranulation flow extends at least to 0.5% of the solar radius below the surface. This research is supported by the SOI-MDI NASA contract NAG5-3077 at Stanford University.