Paleosea-Level Records from Late Quaternary Coral Reef Terraces on Araki Island, Vanuatu; Comparison With Previous Results from Huon Peninsula, Papua New Guinea
Abstract
Robust, proxy-based reconstructions of sea-level change are critical to discerning the processes that drive variability in relative sea level (RSL) in the past, present, and future; however, these reconstructions rely on the ability of statistical models of RSL to accurately constrain the relationship between the proxy and RSL. Corals have complex relationships with sea level and grow in skewed depth distributions below the sea surface. These relationships are poorly-described by traditional methods that assume Gaussian proxy distributions. We develop a new statistical framework to estimate past RSL change based on the modern distributions of RSL proxy elevations in relation to RSL, using corals as an illustrative example. The new statistical framework is hierarchical and consists of 1) data, 2) process, and 3) parameter levels to incorporate nonparametric empirical likelihoods. The data level describes observed sample elevation and geochronological uncertainties from field and laboratory measurements and uses modern depth distributions of individual coral taxa to infer the likelihood of RSL, given the observed proxy elevations. The process level uses the full temporal (or spatio-temporal) covariance to model RSL through time (and space) even where there are no observations. The parameter level dictates prior expectations regarding the correlation structure of RSL. Using Markov chain Monte Carlo (MCMC) sampling, we approximate the posterior distributions on these parameters and RSL, conditioned on the observed data. Through the use of a robust set of synthetic tests, we analyze the sensitivity of the new framework to the quantity and quality of data to which it is applied. We perform these tests on both temporal and spatio-temporal synthetic datasets to assess the ability of the model to estimate RSL and its variability compared to past interpretations of coral proxies. The new framework, while sometimes overestimating uncertainties, performs better than past methods in these tests.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFMPP23G1568R
- Keywords:
-
- 1105 Quaternary geochronology;
- GEOCHRONOLOGYDE: 1120 Isotopic disequilibrium dating;
- GEOCHRONOLOGYDE: 4916 Corals;
- PALEOCEANOGRAPHYDE: 4946 Milankovitch theory;
- PALEOCEANOGRAPHY