Layer boundary determination is important for earth model parameterization in inversion methods. Unfocused logging tools can produce responses that are asymmetric with respect to layer boundaries. This means we cannot extract accurate layer boundaries based on breaks in the slope. We have developed a neural network based approach to pick layer boundaries from unfocused log data by correcting for the shifted response. The neural network approach to layer picking provides a general way to solve the problem of layer boundary determination and avoids the need for a multitude of specialized algorithms for specific situations or tools.
|Original language||English (US)|
|State||Published - 1999|
|Event||1999 Society of Exploration Geophysicists Annual Meeting, SEG 1999 - Houston, United States|
Duration: Oct 31 1999 → Nov 5 1999
|Other||1999 Society of Exploration Geophysicists Annual Meeting, SEG 1999|
|Period||10/31/99 → 11/5/99|
ASJC Scopus subject areas