|
|
|
Objective:Time series analysis methods are developed mostly for uniformly sampled time series. While stratigraphy might be sampled uniformly in the field, once transformed to time this uniformity typically ceases to exist; interpolation must be applied for resampling at a uniform rate. Linear interpolation generally is the most conservative option (although small fraction of variance is typically lost); spline fitting may not appropriate because of overshooting problems (associated with discontinuities or other sharp changes) but is useful for other Earth history applications, e.g., seafloor spreading profiles. Staircase interpolation is needed for stratigraphic series that have been recorded as binary-coded (0,1) stratigraphic series, e.g., marl/limestone couplets.
CONTINUE
|
|
