Datapages, Inc.Print this page

from Outcrop to Subsurface - Comparison between Synthetic Forward Models and Seismic Data of a Channel-Levee System

Szuman, Magdalena 1; Huuse, Mads 1; Kneller, Benjamin C.1
1 Department of Geology & Petroleum Geology, University of Aberdeen, Aberdeen, United Kingdom.

The outcrop used in this study, consisting primarily of the Rosario Formation, is part of San Fernando Channel System and it is exposed over an area of approximately100km2. The main fairway of the channel system runs obliquely to the orientation of the slope; the margins of the channel levee complex are, therefore, non-uniform, with one side confined by the slope itself and the other by the development of an asymmetric levee. The oblique-to-slope orientation and depositional asymmetry of the channel levee system is thought to be the result of large-scale syn-sedimentary deformation.

Detailed outcrop studies on channel evolution and channel body connectivity were used to construct the input geometry for forward modelling. Two models have been created: the first includes a channel complex set and covers 0.8km (vertically) by 8km (horizontally) of stratigraphy of a marine slope channel-levee system. The synthetic seismogram of the section was created using frequencies typically observed on marine seismic data - 30 and 60Hz. The second model covers an individual channel complex within the previously discussed system. The use of very detailed input geometry (at the scale of centimetres) convolved with a high frequency signal (140Hz) provided a unique opportunity to analyze seismic amplitude behaviour bed by bed. All synthetic models have been directly compared to subsurface examples of channel-levee systems and similarities between synthetic seismic and real surveys have been established.

This approach provides the potential to create a set of rules explaining amplitude variation caused by seismically irresolvable geometries. We believe that the knowledge gained by analysis of synthetic seismograms can be directly applied to real seismic data. It is therefore possible to aid the identification of small-scale geometries tuned below the resolution of conventional seismic, thus yielding improved subsurface reservoir prediction.


AAPG Search and Discovery Article #90090©2009 AAPG Annual Convention and Exhibition, Denver, Colorado, June 7-10, 2009