We now compare our predictions with sea-level-change trends inferred from tide-gauge data (Fig. 45.2). We use monthly-average values as supplied by the Permanent Service for Mean Sea Level (PSMSL, e.g. Woodworth & Player, 2003). Details of the tide-gauge stations are listed in Table 45.1. Most stations provide less than 10yr worth of data, the exception being Nuuk (5), which has unfortunately ceased operation. In fact, if the criteria listed by, for example, Douglas (1997) were applied, then none of these time series would be considered usable. However, we will still make use of them to obtain preliminary estimates for present-day sea-level change in south Greenland. We first correct the time series by removing the annual and semi-annual cycles. A linear trend is then fitted to the corrected values, the resulting gradient being the local rate of sea-level change.
Figure 45.2a presents the total predicted present-day sea-level-change rate in south Greenland, combining the results from Figs 45.1c & 45.1d, and a eustatic, or global average, contribution (1.9mm yr-1, Douglas, 1997). Sea level is predicted to be rising all along the southern coast, with millimetres per year spatial variability along the lengths of the larger fjords (hundreds of kilometres). The predicted values for each tide-gauge station are compared with the values inferred from the time series in Fig. 45.2b. The uncertainty range in the predictions is a result of the use of a range of realistic earth-model parameter values, and models of the ice sheets located outside of Greenland (90 to 110% of the nominal ice volume change, Fleming & Lambeck, 2004). Three stations show relatively good agreement between predictions and observations; Sisimiut (3), Nuuk (5) and Ammassalik (7). On the other hand, the values inferred for Ilulissat (1), Aasiaat (2), Maniitsoq (4) and Qaqortoq (6) are quite unrealistic, being of the order of several centimetres per year, the result of the too-short time series that do not accommodate well decadal fluctuations in sea level, leading to large discrepancies between the predicted and observed rates.
Was this article helpful?