Understanding ENSO precursors from a Linear Inverse Modeling Perspective

Abstract:

Predictability of ENSO events depends on the existence and identification of appropriate precursors. Several ENSO precursors have been identified in the literature, both within the Tropics and extratropics. The extra-tropical influences may occur through SST anomalies in the form of “Meridional Modes” in both the northern and southern Hemispheres, as well as SST anomalies in the Western North Pacific. The relative roles of these different types of precursors are, however, still unclear, partly because they also depend upon the background state of the equatorial ocean in ways that are poorly understood at present.

In this study we investigate the efficiency of different aspects of ENSO precursor patterns in generating ENSO. We do this using the well-established Linear Inverse Modeling (LIM) methodology, which objectively identifies optimal initial conditions for ENSO development through a singular vector analysis of the appropriate system evolution operator. We perform this analysis using different observational and climate model datasets. The emphasis is on assessing the importance of local versus remote influences at different stages of ENSO development, and on the interplay between the surface and subsurface variations on that development.