MICROFOSSIL DATA ANALYSES FOR RECONSTRUCTING SEISMIC AND NON-SEISMIC RELATIVE SEA-LEVEL CHANGE

Estimates of elevation derived from microfossil-based transfer functions offer an attractive approach for reconstructing Late Quaternary sea-level histories. To go beyond description of up-core/up-section descriptions of palaeo-water level change requires evaluation of how transfer function methods fit within the overall project design, sampling methodology and hypothesis testing framework. Using diatom and foraminifera data from Alaska, the Pacific Northwest and Australia we show how project design and sampling methodology influence model accuracy and precision, the latter by approximately one order of magnitude. Model accuracy requires validation by at least one measure independent from the transfer function. We question whether published studies adequately consider accuracy and precision of relative sea-level reconstructions.