Geology Reference
In-Depth Information
skipped by the resampling technique raises two
important questions about the discrete sampling
teachnique. First, and most obviously, is the
choice of the resampling interval. Unlike the case
of the Ardmore section where there were prac-
tical physical limitations to making discrete inter-
val observations throughout an outcrop, when
creating a secondary resampled dataset, the size
of the resample interval need only be depend-
ent on computational effi ciency. Ideally, resam-
pling should be done at or below the Nyquist
rate, 0.05 cm in the case of the Cimon del Latemar
section. Conversely, if computational limitations
would make resampling at very high resolution
problematic, an alternative solution is to adjust
the offset, or phase of the resampling interval. In
the example given above it would be possible to
shift the resampling from 0.5, 1.0, 1.5 cm . . . to
0.4, 0.9, 1.4 cm.
By shifting the resampling in this way, it
becomes possible to select an offset that maximizes
inclusion of thin beds in the dataset. Likewise,
each resampled dataset could be analysed using
any of a variety of statistical techniques to evalu-
ate sensitivity to the chosen resampling interval.
Obviously, if comparable results are obtained
at all offsets, the impact of skipping a few beds
would be minimal.
Unquestionably, selection of an appropri-
ate sampling interval is critical to any quanti-
tative analysis of a section sampled using the
discrete facies rank technique. What impact has
exclusion of some of the thinnest occurrences of
Facies 1 had on interpretations of sea-level change
over the Latemar platform? It is reasonable to sup-
pose that at least some aspects of the power and
line spectrogram would be impacted by changes
in the size of the resample interval. While as many
as 24 occurrences of Facies 1 could be skipped by
the technique used by Preto
et al
. (2001), in fact
only eight were with no appreciable change in
the resulting power spectrum when reanalysed
at the Nyquist rate (L. Hinnov, personal com-
munication, 2007).
While the exclusion of some of the occurrences
of supratidal caliche soils from the numerical
analysis of Preto
et al
. (2001) apparently did not
adversely effect their interpretations, the exclusion
of any occurrence of a palaeoenvironmentally sig-
nifi cant subfacies such as the vadose cap and cali-
che soils of the Latemar has the potential to affect
quantitative analysis. Facies 1 can be formed only
when sea-level drops below the platform margin,
making every occurrence, irrespective of thickness,
500
400
300
200
100
Occurrences
of facies1
Very thin
occurrences
Fig. 3.
Stratigraphic distribution of occurrences of the
vadose altered dolostone and caliche subfacies (Facies 1)
in the Cimon del Latemar section (Preto
et al.
, 2001). The
left column illustrates the positions of occurrences of
the facies, while the right column describes the position
of those occurrences that are thinner than or equal to the
scale of the secondary resample spacing (0.5 cm). A total
of 24 of the 162 occurrences of Facies 1 are at or below the
spacing interval.
a third of the observed units (Fig. 3). Interestingly,
however, despite the very small scale resample
spacing of 0.5 cm, nearly 15% of the 162 occur-
rences of the facies fall at or below the resampling
size. Additionally, the stratigraphic distribution
of these very thin caps is not uniform through-
out the section. Rather, 20 of the fi rst 82 and 4
of the last 10 occurrences are at or below the
resolution of the discrete sampling technique
(Fig. 3). Clearly, the observation that very thin
occurrences of Facies 1 are not uniformly distrib-
uted throughout the section is of interest when
considering potential controls of variation in
long-term rates of accommodation space creation
across the Triassic platform.
As the facies most representative of sea-level
lowstands, Facies 1 is critical to any reconstruc-
tion of the temporal record of accommodation
space change across the Latemar platform. The
fact that occurrences of this facies could be
