A new Bayesian approach to the inverse modelling of modern sea level change

Estimates of mean sea level change in the 20th and 21st centuries are important for monitoring the effects of climate change. In particular, there is increasing interest in attributing the relative contributions to observed sea level change both globally and in specific regions. Here we present a new method for obtaining such quantitative inferences from combinations of satellite gravity, satellite altimetry, and tide gauge data. Our approach is based upon a full Bayesian solution to the associated inference problem which incorporates realistic priors on all unknowns along with a comprehensive treatment of observational uncertainties. An essential step within this method is the solution of both the sea level equation and its adjoint, with the latter approach being a new development. As part of this work, open source python libraries are being developed for sea level modelling and the solution of Bayesian inference problems within a function space setting.