Evaluating SODA for Indo-Pacific Ocean decadal climate variability studies

Abstract

Estimates of changes in upper ocean temperature, heat content, and sea level are dependent on the coverage of subsurface observations in space and time. Historically, these data are sparse, which has limited our understanding of ocean climate variability and change mechanisms. Ocean state estimates, which effectively represent a model synthesis and integration of the available observations, including internal observations in the ocean and surface forcing, help to address the inhomogeneity of sparse observations in space and time. Here we evaluate the representativeness of ocean state estimates from the Simple Ocean Data Assimilation Version 2.2.4 (SODA) data for studying Indo-Pacific Ocean decadal temperature and sea level variability over the period . The SODA data are evaluated against independent sea level anomalies from long-record tide gauges at Midway Island and Fremantle, reconstructed sea surface height anomalies, and sea surface height anomalies from TOPEX/Poseidon satellite altimeter observations at the decadal time scale. This study demonstrates that SODA captures the characteristic Interdecadal Pacific Oscillation (IPO) over the upper 200 m, and accurately represents these decadal changes against the independent observations. The SODA-product shows a meridional asymmetry of patterns that connect the western tropical Pacific and the Indian Ocean, apparently in relation to IPO changes. Regional sea level at the Midway Island and Fremantle tide gauges confirm this decadal connection and the relationship with the IPO. We concluded that SODA is potentially a useful tool to examine ocean decadal climate variability across the Indo-Pacific Ocean.

Slow westward movement of salinity anomalies across the tropical South Indian Ocean

“…The EOF3 spatial pattern provides evidence for a connection between the Indonesian Seas and the tropical western NPO and SPO; however the pattern does not penetrate into the tropical SIO beyond about 110°E (Figures e and f). Signals in SSHA related to large‐scale IPO changes have been reported to connect the western tropical Pacific Ocean with the Indian Ocean [ Vargas‐Hernandez et al ., ]. Hence, the signal in salinity anomalies at D20 observed in EOF3 (Figures e and f) is likely to be part of this same signal in SSHA, both related to IPO changes.…”

“…While the Indian Ocean displays a near uniform surface multidecadal warming trend [ Alory et al ., ; Alory and Meyers , ], there is also significant subsurface cooling evidenced in sea surface height (SSH) data (as well as in ocean heat content (OHC), thermocline depth and temperature) and expressed as cooling of the thermocline waters of the tropical South Indian Ocean (SIO) [ Han et al ., ; Alory et al ., ; Trenary and Han , ; Schwarzkopf and Böning , ; Trenary and Han , ]. While this multidecadal signal appears to be the strongest low‐frequency variability in the SIO, there is also evidence of interannual and decadal variability in the dynamics [ Meyers , ; Wijffels and Meyers , ; Cai et al ., ; Alory et al ., ; Feng et al ., ; Nidheesh et al ., ; Trenary and Han , ; Vargas‐Hernandez et al ., ; Han et al ., ], as well as decadal variability of salinity anomalies within the thermocline of the tropical SIO, which is the focus of the present study.…”

“…Low‐frequency variability studies of the tropical SIO have focused on thermal changes associated with OHC, SSH and subsurface temperature, but little research has been undertaken to understand the decadal variability of salinity anomalies and their contribution to steric SSH. For instance, Vargas‐Hernández [] showed that steric SSH had a positive trend whereas the thermosteric SSH (ThSSH) displayed a negative trend for the period 1970–2003 averaged across a zonal band of the SIO (10.25°S–30.25°S, 30.25°E–120.75°E), pointing to the important role that salinity variability plays in the Indian Ocean. Linear trends of steric sea level for short periods (< 30 years, for instance identified in satellite altimeter data from 1993 onward) are quite different from those for longer periods in the Indo‐Pacific Ocean (examples: 1960–1999 in Alory et al .…”

Multidecadal Indian Ocean Variability Linked to the Pacific and Implications for Preconditioning Indian Ocean Dipole Events

“…2). Several other previous studies also related subsurface temperatures/sea surface height/sea level variations in the western Pacific that can be affected by the PDO to (south)eastern Indian Ocean on decadal timescales (e.g., Lee and McPhaden 2008;Schwarzkopf and Böning 2011;Nidheesh et al 2013;Vargas-Hernandez et al 2014), with the relationship strengthening in recent decades (Trenary and Han 2013;Han et al 2014b;Feng et al 2015).…”

“…A concurrent Indian Ocean subsurface cooling in the tropics was associated with more frequent negative Indian Ocean Dipole (IOD) events and a strengthened subtropical cell (Trenary and Han 2008), and a shoaling thermocline (Han et al 2006;Cai et al 2008) in response to changing Pacific wind forcing (Alory et al 2007;Schwarzkopf and Böning 2011). The leading mode of upper-ocean Indo-Pacific temperatures in the Simple Ocean Data Assimilation product was also found to exhibit a long-term trend of surface warming and subsurface cooling at thermocline depth, which Vargas-Hernandez et al (2014, 2015 linked to Pacific modes of climate variability, such as the IPO, North Pacific gyre, and El Niño Modoki. Using sensitivity experiments with an OGCM, Schwarzkopf and Böning (2011) found the Indian Ocean subsurface cooling trend to be reproduced in simulations with observed wind forcing in the Pacific only, while wind stress outside the Pacific was kept at climatology.…”