Sea-level variability in tide-gauge and geological records: An empirical Bayesian analysis (Invited)
Abstract
Sea level varies at a range of temporal and spatial scales, and understanding all its significant sources of variability is crucial to building sea-level rise projections relevant to local decision-making. In the twentieth-century record, sites along the U.S. east coast have exhibited typical year-to-year variability of several centimeters. A faster-than-global increase in sea-level rise in the northeastern United States since about 1990 has led some to hypothesize a 'sea-level rise hot spot' in this region, perhaps driven by a trend in the Atlantic Meridional Overturning Circulation related to anthropogenic climate change [1]. However, such hypotheses must be evaluated in the context of natural variability, as revealed by observational and paleo-records. Bayesian and empirical Bayesian statistical approaches are well suited for assimilating data from diverse sources, such as tide-gauges and peats, with differing data availability and uncertainties, and for identifying regionally covarying patterns within these data. We present empirical Bayesian analyses of twentieth-century tide gauge data [2]. We find that the mid-Atlantic region of the United States has experienced a clear acceleration of sea level relative to the global average since about 1990, but this acceleration does not appear to be unprecedented in the twentieth-century record. The rate and extent of this acceleration instead appears comparable to an acceleration observed in the 1930s and 1940s. Both during the earlier episode of acceleration and today, the effect appears to be significantly positively correlated with the Atlantic Multidecadal Oscillation and likely negatively correlated with the North Atlantic Oscillation [2]. The Holocene and Common Era database of geological sea-level rise proxies [3,4] may allow these relationships to be assessed beyond the span of the direct observational record. At a global scale, similar approaches can be employed to look for the spatial fingerprints of land ice melt [5]. We end by presenting preliminary results from such an analysis. [1] Sallenger et al. (2012), Nat. Clim. Change 2: 884-888. [2] Kopp (in press),Geophys. Res. Lett. [3] Engelhart and Horton (2011), Quat. Sci. Rev. 54: 12-25. [4] Kemp et al. (2011), Proc. Natl. Acad. Sci. 108: 11017-11022. [5] Hay et al. (2013). Proc. Natl. Acad. Sci. 110: 3692-3699.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFMGC33D..04K
- Keywords:
-
- 1641 GLOBAL CHANGE Sea level change;
- 1986 INFORMATICS Statistical methods: Inferential