Summary
Seismic interpretations are often highly subjective and depend on the interpreter’s understanding of the limitations of seismic acquisition and processing as well as the tectonic and depositional environment. Small errors in processing may give rise to features that look like geology. Such processing errors may include improper statics, poor velocity analysis and not adequately removing coherent noise. The result of these errors, if not mistaken as geology, is often classified under the broad category of acquisition footprint.
In this paper we use synthetic seismic data composed of four reflectors to investigate the effects of poor velocity analysis and normal moveout stretch (or migration). We compare the results from the synthetic seismic dataset to a real 3D seismic dataset. We show an attribute interpretation of both datasets and how inaccurate processing can lead to fallacious claims about the geological background.
Introduction
Processing procedures can greatly affect the reliability of conventional interpretation and the utility of seismic attribute interpretation. While seismic modeling is routinely used to calibrate and show the advantages of new processing and imaging algorithms, it is less commonly used to show pitfalls in seismic processing. Very few efforts have been made to explain acquisition footprint and processing generated noise using synthetic models. Hill et al. (1999) discussed acquisition footprint caused by inaccurately picked NMO velocity. Ha (2014) used seismic modeling in an attempt to better understand the response of a fractured granitic basement. He also used elastic modeling to identify coherent seismic noise, such as groundroll. With the insight gained from seismic modeling he was able to better identify and eliminate coherent noise during seismic processing.
Seismic attributes, especially coherence and curvature, often exacerbate the effects of inaccurate processing procedures (Verma et al. 2014; Marfurt and Alves, 2015). Because attributes are popular, particularly among less experienced interpreters, as a method to hasten interpretations this could lead to pitfalls in our geologic model (Marfurt and Alves, 2015).
