1. Introduction
The study of sea level change rate from tidal data is complicated by
the fact that tide gauges are attached to the land and thus their
measurements are affected by vertical land motion caused by natural
or anthropogenic processes unrelated to variations in sea level. In the
traditional approach of many global studies, e.g. Douglas (1997) and
Peltier (2001), only a subset of tide gauges is carefully selected, having
at least 60 years of record length and being far away from tectonically
active areas. Corrections for Global Isostatic Adjustment (GIA) are
then applied to obtain absolute rates of sea level change, which
produce a consistent rising rate at around 1.8 mm/yr for the 20th
century. The main problem with this approach is that the strict criteria
screen out most tide gauges, leaving only around 25 stations for
studying sea level change. Moreover the locations of these qualified
stations are concentrated in the Mediterranean, the North Atlantic
and the East Pacific. No tide gauges from the Indian Ocean and West
Pacific are included and the serious question arises whether this rate
can be truly regarded as global.
More recent studies on the trend and acceleration of global sea
level change mitigate this under-sampling problem by incorporating
as many tide gauges as possible, excluding only those in locations
where vertical movements (caused by tectonic activities or land
subsidence) are known to exist but with rates that cannot be
modelled. In Church and White (2006) the global sea level rise rate
and acceleration were determined from sea level reconstructed from
the amplitude derived from tidal data and the covariance obtained
from empirical orthogonal function analysis (EOF) of altimetry data.
Jevrejeva et al. (2006) and Jevrejeva et al. (2008) used a method based
on Monte Carlo Singular Spectrum Analysis (MC-SSA, quite similar to
EOF), to decompose the original tide gauge time series to detect multidecadal
oscillations of sea level change and its acceleration, which