A probabilistic approach to 21st century regional sea-level projections using RCP and High-end scenarios

Sea-level change is an integrated climate system response due to changes in radiative forcing, anthropogenic land-water use and land-motion. Projecting sea-level at a global and regional scale requires a subset of projections - one for each sea-level component given a particular climate-change scena...

Full description

Bibliographic Details
Published in:Global and Planetary Change
Main Authors: Jackson, Luke P, Jevrejeva, Svetlana
Format: Article in Journal/Newspaper
Language:English
Published: Elsevier BV 2016
Subjects:
Online Access:http://livrepository.liverpool.ac.uk/3064777/
https://doi.org/10.1016/j.gloplacha.2016.10.006
Description
Summary:Sea-level change is an integrated climate system response due to changes in radiative forcing, anthropogenic land-water use and land-motion. Projecting sea-level at a global and regional scale requires a subset of projections - one for each sea-level component given a particular climate-change scenario. We construct relative sea-level projections through the 21st century for RCP 4.5, RCP 8.5 and High-end (RCP 8.5 with increased ice-sheet contribution) scenarios by aggregating spatial projections of individual sea-level components in a probabilistic manner. Most of the global oceans adhere to the projected global average sea level change within 5 cm throughout the century for all scenarios; however coastal regions experience localised effects due to the non-uniform spatial patterns of individual components. This can result in local projections that are 10′s of centimetres different from the global average by 2100. Early in the century, RSL projections are consistent across all scenarios, however from the middle of the century the patterns of RSL for RCP scenarios deviate from the High-end where the contribution from Antarctica dominates. Similarly, the uncertainty in projected sea-level is dominated by an uncertain Antarctic fate. We also explore the effect upon projections of, treating CMIP5 model ensembles as normally distributed when they might not be, correcting CMIP5 model output for internal variability using different polynomials and using different unloading patterns of ice for the Greenland and Antarctic ice sheets.