Quantitative Stratigraphy, Splining and Geologic Time Scales - Geomathematics: Theoretical Foundations, Applications and Future Developments - page 335

Table 9.13 First 28 fossil events in order of ranked optimum sequence for WELL # 18 of dataset
(A): Esso (N) 16(1-1); these events are Last Occurrences (LOs) except for a single Last Common
Occurrence (LCO) and two log markers
# Sequence# Depth (m) Event# Event name

1 1 381 77 Elphidium spp.

2 2 381 228 Cassidulina teretis

3 4 637 1 Neogloboquadrina pachyderma

4 7 381 270 Cibicidoides grossa

5 9 637 4 Globorotalia inflata

6 13 637 23 Sigmoilopsis schlumbergeri

7 14 707 269 Neogloboquadrina atlantica

8 15 726 266 Globorotalia puncticulata

9 17 643 219 Martinotiella cylindrica

10 19 908 285 Caucasina eiongats

11 20 822 91 Diatoms/radiolarians LCO

12 21 1,036 282 Uvigerina ex.gr. semlornata

13 23 1,128 207 NS Log G

14 24 899 125 Neogloboquadrina continuosa

15 25 734 71 Epistomina elegans

16 26 899 15 Globigerina prebulloides

17 28 899 236 G, ex.gr. praescitula zealandica

18 29 908 17 Asterigerina gurichi

19 34 1,395 24 Tumilina alsatica

20 36 1,219 25 Coarse agglutinated spp.

21 37 1,395 97 Cyclammina placenta

22 38 1,395 182 Splrosigmollinella compressa

23 39 1,584 262 Karrerulina horrlda

24 42 1,395 140 Rotaliatina bulimoides

25 43 1,554 261 Haplophragmoides walteri

26 46 1,554 321 Dorothia seigliei

27 47 1,554 289 Adercotryma agterbergi

28 49 1,531 206 NS Log F

Source: Agterberg et al. ( 2013 , Table 5)

depth difference is positive when depth of the event below it is greater and negative
when it is less. Figure 9.22 can be used to illustrate this procedure. On average, the
cluster of points for observed depths dips to the right on this graph. It reflects the fact
that events stratigraphically lower in the optimum sequence tend to have greater
depths. A negative first order depth arises when the depth of an event is less than
that of its neighbor to the right in Fig. 9.22 . Histograms of first-order depth differences
in all wells for a dataset have average value greater than zero. In RASCW Version
20 their frequency distribution can be modeled in various ways. Figure 9.23 (after
Agterberg et al. 2007 )showsnormal Q - Q plots of square root transformed first-order
differences for datasets (A) and (B). The straight lines in Fig. 9.23 represent bilateral
gamma distributions that were fitted as straight lines by least squares excluding data
points on the commonly occurring anomalous upward bulges near the center of each
plot. These anomalies are due to the discrete sampling method used when the wells