Determining the potential risks associated with sea inundation is of
great importance for many coastal communities. The knowledge of rates
of sea level variations is also of fundamental interest for many
academic disciplines. Accurately estimating current rates of
inundation at the regional scale is however complicated by the
interacting contributions of many factors, at various temporal and
spatial scales.


Since the 1990s, satellite altimetry provides a direct measurement of
absolute sea level change, albeit with large uncertainties near
coastlines owing to near shore wave heights. In order to access
variations of relative sea level, the measurement of the absolute
vertical land motion is required as a reference frame and can be
obtained from GPS observations. While tide gauges measure directly the
relative sea level change at local stations, these suffer from a
strong heterogeneity in terms of density of stations as well as the
length of the time series. The signal is also affected by strong
seasonal influences as well as local effects, sometimes of
anthropogenic origin.

In order to better constrain estimates of potential sea level
inundation, we have developed a Bayesian technique for the joint
inversion of tide gauges, land based GPS stations, and satellite
altimetry measurements of sea level.

We will show the results of cross-validation of these differing
observations, in addition to the benefits for constraining coastal sea
level rise obtained through joint inversion of these observations. We
will demonstrate regional case studies in Europe, North America, and
South East Asia. Our results show that by combining estimates of sea
level rise and land uplift, predictions are less influenced by
problematic tide gauges and the estimates of coastal sea level rise
are generally more spatially smooth and perhaps more representative of
broader regional inundation risks.