Synthetic Seismograms With Offset for a Layered Elastic Medium

Abstract

ABSTRACT One of the severest problems facing exploration geophysicists is the interpretation of surface seismic data to estimate the physical properties of potential oil and gas zones. Attempts to do this are usually based dominantly on an inspection of CDP stacked seismic sections, in conjunction with well logs, if available. Existing information on the amplitude and character of the seismic response as a function of source to receiver offset is usually overlooked, mainly because it is not directly interpretable in terms of geological or physical properties. In fact, it may not be feasible to tackle this inversion problem competantly unless the geophysical interpreter has access to a practical solution of the direct elastic problem, namely the theoretical seismic response with offset of an interpreted geological model. Even with today' s computer capabilities, the solution of this direct problem for a complex geologic model is costly. However, the solution can become fast, economical and useful, for a model with horizontal layers only. Illustrative examples, using such a modeling program, are described here and mainly emphasize the possible use of amplitude variations with offset to distinguish a sand-shale sequence from an all shale section. It is believed that this form of elastic modeling may provide a valuable tool for guiding and checking the interpretation of seismic records for stratigraphic oil and gas reservoirs. INTRODUCTION The seismic data used in geophysical interpretation is normally restricted to CDP stacked seismic sections. Sometimes, these are subjected to additional processing to form pseudo acoustic impedance logs which better match corresponding data from well logs in the area. This procedure of using only CDP stacked seismic data overlooks seismic reflection variations with offset that may be of extreme value for stratigraphic interpretation. At this stage of development, we can not readily take measured character variations with offset and unambiguously invert them into the physical properties of the rocks. This is what research and development geophysicists hope to approach in the long term. In the short term, however, it is certain that seismic model experiments can provide an evaluation of the potential of reflectivity variations with offset as a tool for distinguishing between rock types. Optimistically, such experiments could lead to the development of simple characteristics for distinguishing between 'soft' and 'hard' bright spots, and perhaps to methods for estimating the percentage of sand in sand-shale sequences. In fact, two valuable papers on these very topics were presented at the 1982 Annual Meeting of the Society of Exploration Geophysicist (SEG) in Dallas by Ostrander l and Backus et al. This paper presents related experiments using a seismic modeling method particularly adapted to such studies. It is appropriate to note that a similar modeling method was the subject of an excellent verbal presentation by Rosenbaum at the 1971 Annual Meeting of the SEG in Houston, Texas. Unfortunately, this has not been published in manuscript form.