Reconstruction of Local Sea Levels at South West Pacific Islands—A Multiple Linear Regression Approach (1988–2014). Issue 2 (27th February 2018)

• Record Type: Journal Article
• Authors: Kumar, V.; Melet, A.; Meyssignac, B.; Ganachaud, A.; Kessler, W. S.; Singh, A.; Aucan, J.
• Is Part Of: Journal of geophysical research. Volume 123:Issue 2(2018)
• Page Start: 1502
• Page End: 1518
• Publication Date: 2018-02-27
• Subjects: local sea levels -- statistical downscaling; Oceanography -- Periodicals; 551.4605
• DOI: 10.1002/2017JC013053
• Languages: English
• ISSNs: 2169-9275

Abstract

Abstract: Rising sea levels are a critical concern in small island nations. The problem is especially serious in the western south Pacific, where the total sea level rise over the last 60 years has been up to 3 times the global average. In this study, we aim at reconstructing sea levels at selected sites in the region (Suva, Lautoka—Fiji, and Nouméa—New Caledonia) as a multilinear regression (MLR) of atmospheric and oceanic variables. We focus on sea level variability at interannual‐to‐interdecadal time scales, and trend over the 1988–2014 period. Local sea levels are first expressed as a sum of steric and mass changes. Then a dynamical approach is used based on wind stress curl as a proxy for the thermosteric component, as wind stress curl anomalies can modulate the thermocline depth and resultant sea levels via Rossby wave propagation. Statistically significant predictors among wind stress curl, halosteric sea level, zonal/meridional wind stress components, and sea surface temperature are used to construct a MLR model simulating local sea levels. Although we are focusing on the local scale, the global mean sea level needs to be adjusted for. Our reconstructions provide insights on key drivers of sea level variability at the selected sites, showing that while local dynamics and the global signal modulate sea level to a given extent, most of the variance is driven by regional factors. On average, the MLR model is able to reproduce 82% of the variance in island sea level, and could be used to derive local sea level projections via downscaling of climate models. Key Points: Local sea level variability on interannual‐to‐interdecadal time scales in the area is governed mainly by the regional wind field These regional winds could be simulated by climate models, serving as a link between local scale information and low resolution models Mass contribution from melt of glaciers and ice sheets is much smaller relative to the dominant steric component