Models of future sealevel changes

Future sea-level rise has led to concern because of possible impacts to coastal regions, in the form of loss of land through inundation and erosion, increased frequency of storm floods, and saltwater intrusion. Global mean sea-level is a sensitive indicator of climatic change. Global warming will lead to sea-level rise from the thermal expansion of seawater, the melting of alpine glaciers and polar ice sheets. Several studies have reported rates of annual global sea-level rise during the last century ranging between 0.3 and 3mmyr_1. The reliability of these data has, however, been questioned due to problems with data quality and spatial and temporal variations in physical processes involved in sea-level variation, such as wind, ocean currents, river runoff and tectonic movement, of which the latter may introduce a serious source of error. Proxies of palaeo-sea-level changes, geophysical modelling and satellite geodesy may, however, be used to remove isostatic effects. Future sea-level is expected to rise by about 0.5 m by the year 2100, which is 4-7 times faster than present rates. Most global climate models predict a positive net balance in Antarctica because precipitation over the ice sheet may exceed ablation. Because future sea-level projections are quite uncertain, careful monitoring by upgraded tide-gauge networks and satellite geodesy will become important.

According to oxygen isotope ratios in marine sediment cores in relation to astronomical cycles, the present Holocene interglacial period may terminate in a few thousand years (Fig. 6.9). Analyses of solar insolation suggest that the present interglacial may last longer (Berger and Loutre, 1994, 1996), marked by a global fall in temperature and sea-level. Human effects such as the emission of greenhouse gases may, however, lead to the opposite development.




W J"111

,2/v81 / -21&

22 \ 60










Figure 6.9 Long-term climatic variations over the past 400,000 years, and the prediction for the next 60,000 years according to Berger (1988).

400 350

50 0

H = High estimate B = Best estimate L = Low estimate

0 0

Post a comment