We investigate the effects of the ocean function on predictions of the sea-level changes and other geophysical signals due to glacial rebound. To precisely predict these signals, a realistic ocean function including the effects of the palaeotopography, the distribution of ice sheet and meltwater influx is required. The adoption of a precise ocean function is very important in simulating the observables in Hudson Bay for an earth model with a low lower mantle viscosity of ∼10 21 Pa s. In this case, the contribution from water loads can be comparable to that from ice loads. In the Fennoscandian region, however, the predictions are less sensitive to the details of the ocean function, because the width of the Gulf of Bothnia is very small compared with that of Hudson Bay. With an assumption that the ice model is represented by ARC3+ANT4b, we have examined the viscosity structure using relative sea-levels, gravity anomaly and solid surface gravity changes in North America and northern Europe. This study suggests a lower mantle viscosity of greater than 10 22 Pa s and a upper mantle viscosity of (4 ∼ 10) × 10 20 Pa s.

Holocene sea-level observations have been obtained from the Ako Plain in western Japan. In this coastal area, the Holocene crustal movements have been evaluated, except the crustal response due to the last deglaciation, by comparing observations and theoretical predictions of relative sea-level (RSL) variations. Analyses of diatom assemblages and sedimentary sulphur in core sediments were used along with radiocarbon dates to derive the RSL variations. A crustal movement rate between +0.2 mm and -0.2 mm per year, corrected for the prediction, fits well with the observed RSL index points. Owing to the constraint of the reconstructed palaeo-mean sea level (PMSL) at 7300 cal. BP, the coast of the Ako Plain may have the best estimate of a tectonic subsidence rate of 0-0.2 mm/yr. The tectonic uplift rates along the tectonically active coast of western Kobe were derived to be 0.3 -0.7 mm/yr and 0.11 -0.45 mm/yr for Tarumi and Tamatsu, respectively, relative to Ako over the period concerned. The relative uplift along the traverse from Ako to western Kobe is primarily the result of the crustal movement resulting from the active faulting of the Rokko-Awaji fault system (RFS).