Geology Reference
In-Depth Information
Evaluating validity and reliability in high-resolution stratigraphic
analysis
CARL N. DRUMMOND and LAILAH A. MARLOW
Department of Geosciences and Offi ce of Research and External Support, Indiana University-Purdue University
Fort Wayne, Fort Wayne, IN 46805-1499 260 481 5750, USA (E-mail: drummond@ipfw.edu)
ABSTRACT
The analysis of stratigraphic sections has always been conducted through an
integration of quantitative and qualitative observations. Recently, new techniques
have been put forward as mechanisms for extracting high-resolution quantitative data
from outcrops for the purpose of assessing controls on accommodation-space change.
Evaluation of the validity and reliability of these analytical techniques provides an
important framework for future implementation. First, the discrete facies rank tech-
nique is employed by stratigraphers as a mechanism for the construction of a proxy
sea-level curve. Exclusion of highly diagnostic but thin or infrequently occurring
facies, as driven by sampling interval size, will potentially have an impact on the
results of analysis of stratigraphic organization. The probability of inclusion of a facies
in a dataset of equally spaced lithological observations is controlled by the thicknesses
of the occurrences of the lithology as well as its recurrence frequency in the section.
Second, the grey-scale analysis of outcrop photographs is used as a quantitative
proxy for lithological variation. Evaluation of this technique indicates that it is highly
dependent upon the geometric relationships between the outcrop and the focal plane
of the camera. Geometric foreshortening can result in quantitative distortion of the
grey-scale series, and thus has impacts on interpretations drawn from the analysis of
the data series. Additionally, scale-dependent averaging of grey-scale values across a
row of pixels can result in quantifi able distortion in the spectral characteristics of the
data series. While both the discrete facies rank technique and grey-scale analysis of
outcrop photographs can provide useful mechanisms for the collection of uniformly
spaced stratigraphic data, great care must be taken in how these techniques are used
and how results from the analysis of resultant data are interpreted.
Keywords
Cyclicity, Milankovitch, photographic interpretation, Fourier analysis.
INTRODUCTION
in a variety of ways to evaluate spatio-temporal
variation in depths, durations and modes of
deposition. Stratigraphic analysis has likewise
been complicated by the several orders of mag-
nitude difference between the scales and reso-
lutions of outcrop and subsurface observations.
Subsurface methods return continuous records of
geophysical parameters that are readily evaluated
by a wide range of statistical methods. Conversely,
fi eld-based observations are commonly more
discrete and more qualitative in their character.
Recently, several techniques for converting some
of the complex qualitative aspects of outcrop
observation into a continuous data series suitable
for spectral analysis have been advanced. It is
the purpose of this paper to describe and evalu-
ate two analytical techniques that have received
use in the quantitative stratigraphic literature, with
The analysis of stratigraphic successions has
typically been achieved through application of an
array of quantitative and qualitative techniques.
Bed thickness, grain size, dip direction of cross-
bedding and many other types of stratigraphic and
sedimentological observations are intrinsically
quantitative in their character. Measurements are
made within the limitations of accuracy and pre-
cision and the resultant data are evaluated using a
wide range of numerical techniques. Conversely,
selection of the number and type of lithostrati-
graphic elements observed in a succession and
their palaeoenvironmental relationships are sub-
jective and qualitative measures of the nature of
the stratigraphic record. Taken together, qualita-
tive and quantitative analyses have been utilized
