Probabilistic inversion of sea level, tectonics, and coral reef morphogenesis

Coastal landforms constitute a worldwide archive that intricately records past sea level, tectonics, and geomorphic processes, yet unraveling the complex interplay of these processes driving the morphogenesis of the shorelines remains a challenge. Albeit forward landscape evolution modeling has proven successful in deciphering either sea-level, tectonics, or geomorphic processes independently, this strategy rests upon hypotheses regarding the other two aspects. We circumvent this limitation by inverting coral reef terrace morphologies using the new model RoSETTA (Resolving Sea level, Ecosystems and Tectonics from Terrace Analysis), which produces probabilistic reconstructions of past sea level, tectonic vertical motion, erosion processes, and reef growth rates from coastal landforms. We first benchmark RoSETTA on synthetic terraces sequences and then on the canonical landforms of Sumba island, Indonesia. Beyond reconstructions, our probabilistic approach allows to highlight periods of time and sea level that can confidently be resolved from the geomorphological archive, and conversely identifies periods where the archive is insufficient to be conclusive. A successful remedy is to jointly invert multiple profiles using RoSETTA, improving the confidence of the results. We provide here a probabilistic reconstruction of continuous relative sea-level history and rates of change over the Late Quaternary. Periods of time with sea-level rise rates analogous to present day and projected rises are identified, and coral reef responses to such conditions are quantified over millennial timescales, indicating that Sumba-type coral reefs could hardly keep-up with IPCC-predicted rates of sea-level rise for the next century.