Denovan Chauveau , Anne-Morwenn Pastier, Gino de Gelder , Laurent Husson, Christine Authemayou, Kevin Pedoja, Sri Yudawati Cahyarini

The morphology of coastal sequences provides fundamental observations to unravel past sea level (SL) variations. For that purpose, converting morphometric observations into a SL datum requires understanding their morphogenesis. The long-lasting sequence of coral reef terraces (CRTs) at Cape Laundi (Sumba Island, Indonesia) could serve as a benchmark. Yet, it epitomizes a pitfall that challenges the ultimate goal: the overall chronology of its development remains poorly constrained. The polycyclic nature of the terraces, involving marine erosion and reoccupation of old coral colonies by more recent ones hinders any clear assignment of Marine Isotope Stages (MIS) to specific terraces, in particular the reference datum corresponding to the last Interglacial maximum (i.e., MIS 5e). Thus, to overcome these obstacles, we numerically model the genesis of the sequence, testing a range of eustatic SL reconstructions and uplift rates, as well as exploring the parameter space to address reef growth, erosion, and sedimentation. A total of 625 model runs allowed us to improve the morpho-chronological constraints of the coastal sequence and, more particularly, to explain the morphogenesis of the several CRTs associated with MIS 5e. Our results suggest that the lowermost main terrace was first constructed during the marine transgression of MIS 5e and was later reshaped during the marine regression of MIS 5e, as well as during the MIS 5c and MIS 5a highstands. Finally, we discuss the general morphology of the sequence and the implications it may have on SL reconstructions. At Cape Laundi, as elsewhere, we emphasize the necessity to address the development of CRT sequences with a dynamic approach, i.e., considering that a CRT is a landform built continuously throughout the history of SL oscillations, and not simply during a singular SL maximum.