Imaging the lower slope, offshore Nicaragua and Costa Rica using a new residual migration velocity analysis technique in the space-offset domain

There is a dramatic variation in geochemical sediment tracer signal along the Central American volcanic arc. Two contradicting theories, one supporting sediment accretion and the other subduction erosion, have been suggested as possible explanations for these variations. My goal in this dissertation is to use seismic images of the lower slope off Nicaragua and Costa Rica to study the influence of subducting plate structure on sediment dynamics. I am particularly interested in documenting the efficiency of sediment subduction, its along strike variation, and to see if it corresponds with the geochemical anomalies. Although high quality seismic data was acquired offshore Nicaragua, it remained difficult to image the lower slope in detail with conventional processing techniques. Several characteristics of this geologic environment pose seismic imaging problems: the area is heavily faulted, adjacent reflection boundaries have contrasting dips, and the velocity structure is complex. In this environment the common midpoint gathers do not represent a collection of true common subsurface reflection points. To get a clearer image, I needed to apply pre-stack depth migration (PSDM) techniques. An accurate velocity model is required to get a good PSDM image. Therefore one of the most important aspects of PSDM is velocity analysis. Over the last few decades, residual migration velocity analysis (RMVA) has been an area of active research. Previous work on RMVA in the depth-offset domain required top down layer stripping migration in order to derive the interval velocities directly, hence making it very computationally intensive. Here I propose a new technique in which for each common image gather (CIG) we first create a table of offset-ray parameters-depth (x-p-z) using a local 1D assumption. Then I calculate the residual migration depth corrections in the p-z domain and finally map these depth corrections back to the x-z domain using the x-p-z table. Since I calculate the residual migration depth corrections in the p-z domain, the interval velocities are derived directly by top down residual migration. Hence I do not have to explicitly do the layer stripping migration followed by residual normal moveout to get the interval velocities. The velocities generated by using the above RMVA technique produce interpretable depth images of the lower slope off Nicaragua and Costa Rica. From these images I interpret partial sediment accretion off Costa Rica and likely total sediment subduction off Nicaragua, and therefore provide a geophysical evidence for the observed geochemical disparity along this margin.