Spatio-temporal Variability of El Niño Southern Oscillation from Geodetic Satellites and Model Data

The global sea level varies primarily by two causes: thermal expansion and mass transports. They reflect the responses of the oceans to global climatic changes, but their relative share in the total sea-level variation (SLV) as a function of space-time is a very complex phenomenon to be continuously monitored. We examine several geophysical data sets (for SLV, ocean temperature, surface currents, mixed-layer depth), using the method of C/EOF (Complex/Empirical Orthogonal Function) to study the space-time variability and propagation of the El Niño Southern Oscillation (ENSO) in the tropical oceans. The SLV from satellite altimetry reveals that from 1993 to 2009 the first C/EOF mode’s time series has high correspondence to the SOI and Nino3.4 index, whereas the time variable gravity obtained from the GRACE satellite data indicates weaker mass migration signals than SLV over the ENSO region, although larger uncertainty is expected from GRACE data over low latitudes. Thus we can characterize the steric vs. mass-induced SLV for ENSO. We also examine the in-situ and OGCM output (such as Ishii & Kimoto (2009) and ECCO-2), e.g. salinity and temperature profiles, and identify the depth anomaly of the ocean mixed-layer of the ENSO region and its variability over the past few decades, showing strong ENSO signals but a general underestimation in the OGCMs.