 |
| Titel |
Uncertainty quantification of Antarctic contribution to sea-level rise using the fast Elementary Thermomechanical Ice Sheet (f.ETISh) model |
| VerfasserIn |
Kevin Bulthuis, Maarten Arnst, Frank Pattyn, Lionel Favier |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250150419
|
| Publikation (Nr.) |
 EGU/EGU2017-14880.pdf |
|
|
|
|
|
| Zusammenfassung |
| Uncertainties in sea-level rise projections are mostly due to uncertainties in Antarctic
ice-sheet predictions (IPCC AR5 report, 2013), because key parameters related to the current
state of the Antarctic ice sheet (e.g. sub-ice-shelf melting) and future climate forcing are
poorly constrained. Here, we propose to improve the predictions of Antarctic ice-sheet
behaviour using new uncertainty quantification methods.
As opposed to ensemble modelling (Bindschadler et al., 2013) which provides a rather
limited view on input and output dispersion, new stochastic methods (Le Maître and Knio,
2010) can provide deeper insight into the impact of uncertainties on complex system
behaviour. Such stochastic methods usually begin with deducing a probabilistic description of
input parameter uncertainties from the available data. Then, the impact of these input
parameter uncertainties on output quantities is assessed by estimating the probability
distribution of the outputs by means of uncertainty propagation methods such as Monte Carlo
methods or stochastic expansion methods. The use of such uncertainty propagation methods
in glaciology may be computationally costly because of the high computational complexity
of ice-sheet models. This challenge emphasises the importance of developing reliable and
computationally efficient ice-sheet models such as the f.ETISh ice-sheet model (Pattyn,
2015), a new fast thermomechanical coupled ice sheet/ice shelf model capable of
handling complex and critical processes such as the marine ice-sheet instability
mechanism.
Here, we apply these methods to investigate the role of uncertainties in sub-ice-shelf
melting, calving rates and climate projections in assessing Antarctic contribution to sea-level
rise for the next centuries using the f.ETISh model. We detail the methods and show results
that provide nominal values and uncertainty bounds for future sea-level rise as a reflection of
the impact of the input parameter uncertainties under consideration, as well as a ranking of
the input parameter uncertainties in the order of the significance of their contribution to
uncertainty in future sea-level rise. In addition, we discuss how limitations posed by the
available information (poorly constrained data) pose challenges that motivate our current
research. |
|
|
|
|
|
|