Figure 4

Illustration of sea-level change with observations and predictions for the Bonaparte Gulf and other MIS 2 sea-level records. Dashed lines correspond to ice volume equivalent sea level, and black lines to predictions at elastic lithosphere of 70 km, upper mantle viscosity of 4 × 10²⁰ Pa s, and lower mantle viscosity of 5 × 10²² Pa s. (a) RSL data in the Bonaparte Gulf based on the depth-transect approach. The isostatic differences among cores are corrected to the sea-level equivalent value at KH11-1-GC06. Red symbols indicate data from this study. Whites correspond to data from previous work^13,14. The light blue bar is the duration of sea-level minimum. The dark blue bar is the duration of sea-level plateau. The depth transect indicates that marginal marine facies^13,14 occur within +25 m water depth and shallow marine^13,14 facies occur over +25 m water depth. (b) Reconstruction limited 22 to 18 cal kyr BP. Symbols same as panel (a). The depth transect indicates that lagoonal/estuarine and intertidal facies occur within +15 m water depth using points from KH11-1-GC03 (point X) and GC06 (point Y) at ca. 20.5 cal kyr BP. (c) Reconstruction limited 12 to 18 cal kyr BP. The point α cannot be explained by a single sea-level curve deduce from point β if paleo-water depth uncertainty is +5 m, suggesting that the uncertainty of the marginal marine environment (inverted triangles) should be revised to +25 m. The brackish environment is observed at ca. 13 cal kyr BP (point γ)^13,14. However, this core was not reevaluated in ref. ¹⁴. Other sea-level data from corals⁶ and sediments⁸ show the sea-level position above −80 m, suggesting that the paleo-water depth of this brackish facies is underestimated. (d) RSL reconstruction from the Huon Peninsula^10,11,12. (e) RSL reconstruction from the Sunda Shelf^8,9. Gray shades show radiocarbon ages of acid-insoluble and leachable organic matter in the same horizon. Uncertainties of age derive from the effect of old carbon³⁰. (f) RSL reconstruction from Barbados^6,7.