Abstract

Sequence stratigraphy has the potential to provide a consistent method for integrating data, correlating strata, defining stratigraphic evolution, and generating quantifiable predictions. However, the consistent application requires a precise definition of concepts, stratigraphic units, bounding surfaces, and workflow. Currently no single generally accepted approach to sequence stratigraphic analysis exists, nor are there any robust tests of models and methods. Applying conventional sequence stratigraphic analysis to strata from an analog laboratory experiment (eXperimental EarthScape02, XES02) with known boundary conditions and chronology provides some initial robust testing of the models and methods. Despite stratigraphic architectures apparently consistent with those expected within the sequence stratigraphic paradigm, blind-test applications yield: 1) deducted erroneous base-level curves, 2) systems-tract classification mismatches, 3) disconnected systems-tracts type and actual base level, 4) time-transgressive basin-floor fans, and 5) missing systems tracts. Stratigraphic forward models using base-level curves derived from Wheeler diagrams cannot match the timing, redeposited-sediment volume, and depositional environments observed in the XES02 experiment. These mismatches result from common Wheeler diagram construction practice, producing poorly resolved base-level minima timing and base-level fall durations, hence inaccurate fall rates. Consequently, reconstructions of controlling factors based on stratal architectures remain uncertain, making predictions similarly uncertain. A reasonable path forward is to properly acknowledge these uncertainties while performing stratigraphic analysis and to address them through multiple scenario analysis and modeling.