Quantitative Stratigraphy, Splining and Geologic Time Scales - Geomathematics: Theoretical Foundations, Applications and Future Developments - page 341

Fig. 9.25 Schematic illustration of the event-depth curve from RASC output for a well (After
Agterberg et al. 1985 ); (a) RASC distance has been replaced by age ( y ) and a cubic smoothing spline
is used to express y as a function of x representing relative event level in a well; bar in x denotes use of
regular sampling interval for x; smoothing factor (SF), whichwas selectedbefore the curve-fitting using
one age per level, is slightly smaller than standard deviation (SD) for all original values; (b) spline curve
g ( x ) is fitted to express depth as a function of event level; SF SD is some very small value; (c) spline

curve g ( x ) recoded as set of values for x at regular interval of z ;(d) curve passing through set of values of
y at regular interval of z obtained by combining spline curve of Fig. 9.25a with that of Fig. 9.25c ;
(e) spline curve fitted to values of Fig. 9.25d usingnewsmoothingfactorSF;(f) standard deviation SD
is computed after curve-fitting using one age per level (Source: Agterberg 1990 , Fig. 9.6)

¼

Ordinary scaling in RASC is based on the assumption that all events have normal
frequency distributions with equal variance along the interval scale. In general,
different events have different frequency distributions. In the analysis of variance
option of RASC, approximate estimates of these frequency distributions can be