20th Century Trivariate Indian Ocean Regional Sea Level Reconstruction

The low-lying densely populated coastal zones of the countries that border the Indian Ocean (IO) are home to over 350 million people that are at risk due to sea level (SL) rise. Accurate regional SL rise estimates are vital to help policymakers effectively aide planning and adaption strategies to help limit socio-economic impacts associated with SL rise. At decadal time scales, internal climate variability is capable of influencing regional SL and to help quantify this we require long-term SL datasets. Reconstructions, re-analyses and products based on in-situ data only, are some of the methods available to fill spatial and temporal gaps left by tide gauges in the IO. However, inconsistencies identified across these products motivate us to improve the regional estimates of IO SL variability. In this study, we build upon an existing SL reconstruction technique by including sea level pressure as a third variable (trivariate) which along with sea level and sea surface temperature helps to greatly improve reconstructed SL estimates. This technique is particularly advantageous for reconstructing IO SL during the first half of the 20^th century which is characterized by a poor spatio-temporal TG coverage. Using cyclostationary empirical orthogonal functions (CSEOFs) we generate IO regional basis functions that have been observed to differ from those computed globally. Using this new regionally reconstructed dataset that spans 1900 to 2013, we seek to improve the understanding of the interdecadal modulation of the Indian Ocean Basin mode (IOB), Indian Ocean Dipole (IOD), Subtropical Dipole Pattern (SDP) and their relative contributions to the 20th century IO regional SL variability. Results from this study ultimately aim to identify the impact of anthropogenic forcing on regional IO SL by removing the fraction of variability associated with internal climate variability.