Conventional or continuous velocity logging generally fail to solve the problem of velocities in the whole area of a seismic survey. Surface measurements often bring a large amount of additional information: effect of shallow layers, change of facies, correlations. The authors present a theoretical view on the physical meaning of interval velocities. After a short synthesis of the diverse formulae formerly proposed (Faust, West), the practical manner of using these formulae and the accurate interpretation obtained with the help of geological and geophysical data are shown. Several characteristical examples are finally explained and commented on.

The Karroo of Madagascar is a thick continental deposit of sandstone with some shale, but no characteristic beds. Where it outcrops, velocities have been measured through bore hole surveys. It seems that they depend primarily on the age, but long refraction soundings show a progressive increase of the velocity with depth and a weak differentiation of the marker beds. These results may possibly be brought in accordance by discriminating between the present position and the deepest position of a bed through its geologic history. Thus we must admit very important erosions which, according to regional geologic data, seem possible. If we make the appropriate correction due to this distinction, all the velocity measurements fall near the curve proposed by Faust for an average shale and sand section.

No abstract

No abstract

No abstract