GRAPHIC CORRELATION

Graphic correlation was developed by Alan Shaw during the late 1950s with the aim to solve stratigraphic problems for which traditional stratigraphy could not provide a solution due to limited resolution. The methodology of graphic correlation is rather simple. However, utilization of its full potential requires a large archival database of stratigraphic ranges composited from hundreds to thousands of sections. Construction of such a Composite Standard Database involves a long-term, dedicated collection effort. Graphic correlation analysis begins by selecting the most complete available section called the standard reference section (Olin Mann and Lane, 1995). Data from a second best section is then composited via graphic correlation to the reference section. The resulting integration represents the incipient composite standard (Figure 1).

The extent of faunal ranges in the composite standard is expressed in units of rock thickness from the standard reference section. In practice, this section can be subdivided into smaller, arbitrary intervals called Composite Standard Units. Although, Composite Standard Units have a lineary distribution relative to the thickness of the standard reference section, they do not represent equal amounts of geologic time due to variation in the in rock accumulation rate that characterizes that section. However, a tie between the composite standard and geological time can be accomplished via integration of sections with absolute time datums such as radiometric dates, paleomagnetic units and orbital cyclicity. (Figure 2).

Graphic correlation correlates two intervals via stratigraphic datums that the two sections have in common. Additional sections can then be composited to the resultant stratigraphic reference section by graphic comparison of datums that these sections have in common. This progressively growing composite data forms the Composite Standard Database. As more and more sections are integrated, the composited ranges in the database eventually approach the true range from the first to last stratigraphic appearance of the datums either locally or globally depending upon the geographic distribution of the chosen sections. A relatively mature database requires from 30 to 40 sections for a relatively small area to thousands of sections for a global system .

In general, datums used in graphic correlation consist of fossil first and last stratigraphic occurrences. However, any potentially isochronous stratigraphic event can be defined as a graphic datum, from absolute age events (e.g., radiometric date) to regional key beds such as bentonites or potential sequence boundaries defined by well log and/or seismic data.

When graphically correlating sections, the paleontologist can be presented with a large number of possible correlation datums. He/she must use their judgement to determine which of these datums represent true first or last stratigraphic occurrences and which do not, the latter indicating truncation of their ranges due to factors such as erosion or facies exclusion. A line of correlation between the sections is drawn by connecting the datums deemed to represent complete stratigraphic distributions. The successfully drawn line of correlation is a time/depth plot, a one-to-one correlation between the section being composited and the Composite Standard Database. The line can be used by the graphic correlation software to adjust stratigraphic datums that are deemed to be immature in the Composite Standard Database via comparison to the newly composited section; i.e., first stratigraphic appearance datums that occur to the left of the line and last stratigraphic appearance datums that occur to the right of line will be extended backwards and forwards in time to a point on the line of correlation.

Figure 2

As Carney and Pierce (1995) stressed, graphic correlation is not a “Black Box” to which one introduces data and automatically receives a stratigraphic correlation. Rather, graphic correlation is a tool that a paleontologists can use to make informed decisions for the biostratigraphic interpretation of a given section (outcrop or well) based upon a system designed for compiling, storing, organizing, and retrieving large volumes of data. All factors considered in traditional biostratigraphy must also be taken into account for graphic correlation. Thus, a full understanding of the local stratigraphy is important to have for a successful graphic correlation; e.g., factors such as faulting, unconformities, microfossil reworking, facies changes, diagenesis, and fossil taphonomy and preservation must be considered.

Figure 3