Environmental Research Letters (Jan 2013)

Evaluating the ability of process based models to project sea-level change

  • John A Church,
  • Didier Monselesan,
  • Jonathan M Gregory,
  • Ben Marzeion

DOI
https://doi.org/10.1088/1748-9326/8/1/014051
Journal volume & issue
Vol. 8, no. 1
p. 014051

Abstract

Read online

We evaluate the ability of process based models to reproduce observed global mean sea-level change. When the models are forced by changes in natural and anthropogenic radiative forcing of the climate system and anthropogenic changes in land-water storage, the average of the modelled sea-level change for the periods 1900–2010, 1961–2010 and 1990–2010 is about 80%, 85% and 90% of the observed rise. The modelled rate of rise is over 1 mm yr ^−1 prior to 1950, decreases to less than 0.5 mm yr ^−1 in the 1960s, and increases to 3 mm yr ^−1 by 2000. When observed regional climate changes are used to drive a glacier model and an allowance is included for an ongoing adjustment of the ice sheets, the modelled sea-level rise is about 2 mm yr ^−1 prior to 1950, similar to the observations. The model results encompass the observed rise and the model average is within 20% of the observations, about 10% when the observed ice sheet contributions since 1993 are added, increasing confidence in future projections for the 21st century. The increased rate of rise since 1990 is not part of a natural cycle but a direct response to increased radiative forcing (both anthropogenic and natural), which will continue to grow with ongoing greenhouse gas emissions.

Keywords