Using forward modelling to test hypotheses on age-depth relationships in marine carbonate records

The goal of age-depth models is to quantify the uncertainties of age tie-points and account for error propagation of these uncertainties. However, age-model inference hardly uses sedimentological information. Carbonates, in theory, should be highly resolved time archives: their mineralogical and skeletal composition is the product of changes in the environment, and they are commonly autochthonous. Radically different concepts of how time is recorded in carbonates coexist in the sedimentological community. Existing paradigms alternatively attribute this spatiotemporal distribution to astronomical forcing, autogenic processes, or stochastic variation. Rarely are these alternatives subject to systematic testing and, when tested, they are often assumed to be mutually exclusive.
Forward models allow systematic testing of hypotheses on the preservation of time in strata under different driving mechanisms and on the distinguishability of these mechanisms. We present an ensemble of stratigraphic architectures generated with such model under alternative hypotheses on the mechanism of sediment accumulation. A classification of these architectures allows us to answer the questions: can astronomical forcing be reliably identified in carbonate stratigraphic record? When will autogenic patterns be confounded with astronomical forcing? By allowing multiple drivers and their combinations, we show how age-depth models can be used as scientific hypotheses and tested against empirical data.