Abstract Reflection seismology is the science of examining the earth’s interior through the analysis of mechanical waves. Seismic waves are bent, reflected, refracted, diffracted, and scattered. The emitted signal is weakened by these effects as well as geometric spreading and attenuation. Seismic waves are generated by controlled sources and travel through fluids, solids, and porous solids. In this chapter, we consider those properties of waves that do not depend on the kind of material supporting the wave propagation. For example, Snell’s law is a general property of wave motion but differs in detail between reflection of acoustic and elastic waves. In this chapter, when considering a general property such as Snell’s law, we will discuss the scalar (acoustic) wave case.

Abstract The physical properties of earth materials that control mechanical wave propagation are numerous. A short list would include mineralogy, lithology, porosity, pore fluid, fractures, density, and permeability, as well as large-scale structure and stratigraphy. All of these properties can and do vary from point to point in the earth. To make progress in seismic simulation and processing, a mathematical model is used to approximate wave propagation and understand phenomena. This model can assume the earth is a fluid, a solid, or a porous solid. As with many other topics, we develop the theory of seismic waves in progression from the simplest useful setting to complex models that exhibit most of the wave effects seen in real data.

Abstract The fundamental experiment in reflection seismology is a single source shooting into many receivers resulting in a collection of field, or prestack, seismic traces containing certain events. In this chapter, we describe and attempt to understand the different kinds of events likely to be seen in field data. By understand, we mean the ability to recognize events by their traveltime characteristics, amplitude variations, and relationship to other events. Furthermore, we want to understand the subsurface conditions that give rise to observed events.

Abstract The previous chapter developed an understanding of seismic events that assumed an acoustic earth model. This is a convenient way of dealing with seismic P-wave data without addressing more complicated elastic phenomena. So, in effect, the fluid medium of the last chapter was a generic proxy used to describe P-waves traveling in the earth. However, there are three actual fluids whose detailed properties are of interest: natural gas, petroleum, and water. The goal is to have a predictive model to calculate the density, bulk modulus, and sound wave speed for these fluids individually or as a fluid mixture. This is important for amplitude versus offset, time-lapse seismic, and fluid substitution work.

Abstract An acoustic model of the earth is a useful approximation. However, acoustic seismic simulation will not predict elastic effects (mode-converted waves, nonacoustic reflection amplitudes, etc.), and acoustic processing will deal with them improperly. Despite these limitations, industry practice at the current time is almost entirely acoustic seismic imaging, although there are circumstances where elastic effects cannot be ignored or yield fundamentally new information. In any case, it is important to understand the basic physics of elastic wave propagation.

Abstract In a broad sense, sedimentary rocks can be subdivided into siliciclastic (conglomerate, sandstone, shale, silt, mudstone) and carbonates (limestone and dolomite). Each group has detailed classification schemes that depend on mineralogy, rock fabric, grain-size distribution, and other small-scale properties. In petroleum seismology, we probe such rocks with 50–200 m wavelengths, and our waves are influenced only by the average properties over this kind of distance.

Abstract Acquisition geometry for 2D seismic data contains many elements common to 3D acquisition. We will review 2D acquisition by going, in detail, through items in the geometry section of the line header for a specific west Texas seismic line shot for Unocal in 1990. Where it will clarify matters, throughout the description the original item name will be placed in parentheses (e.g., traces/record) for reference. In this chapter, we deviate from the preferred metric system because the line header lengths are in English units (1 ft = 0.3048 m). However, the concepts given here are equally valid for a 2D seismic line shot with metric intervals. Rather than go pedantically through the items from first to last, we subdivide them into topics.

Abstract Economics of 3D seismic surveys are affected by some geophysical decisions, such as sampling density and survey area, as well as numerous nongeophysical factors. These include market prices for petroleum and natural gas, leasing situations, drilling costs, etc. The detailed design of the survey (source and receiver locations, fold map, etc.) is done only after economic feasibility has been determined.

Abstract Major items that drive seismic survey design can be divided into three categories. Operational issues include physical access to the acquisition area, supply logistics, cultural interference and noise, and interference from other surveys. Economic issues involve the survey budget, field and prospect economics, and seismic crew costs and availability. Important geophysical issues concern us in this chapter. The parameter choices discussed here are common to 2D and 3D data and have a strong influence on data quality.

Abstract The development of 3D seismic on an industrial scale required parallel advances in recording systems, computers, and processing technology. The first large-scale 3D seismic survey was undertaken in 1971 by Geophysical Service Incorporated on behalf of an industry consortium. By the mid-1980s, 3D seismic was in wide use. Today it is the standard form of petroleum seismic data worldwide. Two-dimensional seismic data continue to be acquired but represent an ever smaller segment of the seismic industry, primarily as a reconnaissance tool in frontier exploration areas. Rather than track the history of 3D seismic acquisition, we will concentrate on a few fundamental geometries that lead naturally to template shooting, which is the most common method in use today.

Abstract There are many trade-offs in the design of any seismic survey. For 2D seismic, there are so few parameters that experience can effectively be used to achieve an optimum design. In 3D the situation is different. Marine 3D is strongly constrained by hardware arrangements, including number and length of streamers, and the fact that the source is in a fixed position relative to the live receiver spread. But land 3D has complete flexibility with respect to design, and this means that determining an optimum shooting arrangement is more challenging.

Abstract Marine seismic surveys are of generally higher quality and have better subsurface accuracy than land data. The reasons include a lack of a weathering layer, no coupling problem between the medium (water) and the receiver, and less ambiguity for pressure measurements than ground motion observations. However, marine data also have problems not found in land data, including water-bottom multiples, source-receiver ghosts, and more complicated positioning issues.

Abstract The term 4D is commonly used in seismic work, as well as in some medical and computational fluid dynamics applications. In the seismic case, this can be a confusing and misleading term, and our purpose here is to clarify the role of data dimensions and components and move toward a consistent and sustainable way of describing data.

Abstract Raw seismic data bear no resemblance to features inside the earth. It is primarily an expression of the experimental details of how the data were acquired. Seismic data processing draws on full knowledge of wave propagation and acquisition geometry to create geologically meaningful images of the subsurface.

Abstract Geophysics is tightly linked to computing technology and power, yet many geophysicists will never log on to a supercomputer. They live in a world populated by workstations and desktop computers. The work is interpretation, log analysis, synthetic seismograms, and report presentation preparation. They get along just fine without supercomputers. But the geophysicist who does not understand computers is at the mercy of those who do.

Abstract Long before the advent of prestack migration, a processing flow was developed to sequentially remove residual wave propagation and acquisition geometry effects in seismic data. Where prestack migration bundles all this functionality in one grand program, the traditional approach isolates each effect for analysis and optimum removal. The common midpoint (CMP) stack approach is still the industry standard unless subsurface conditions dictate the use of prestack migration or prestack interpretation is planned, such as amplitude variation with offset (AVO) work or azimuthal anisotropy estimation for fracture mapping.

Abstract Migration is a process that compensates for effects introduced by wave propagation and acquisition geometry. It is a subject of great breadth and challenging mathematical complexity. For an advanced, yet very readable, account of migration, the reader is referred to Claerbout (1985), and a comprehensive treatment is available in Yilmaz (2002). Here we develop the conceptual framework of migration and present theoretical results only when they serve to reinforce the discussion.