Internal climate variability in a state space of statistical moments: ENSO and its asymmetry
Abstract
Intrinsic climate variability (ICV) is characterized by the first and second moments of a climate variable and subjected to system analysis in (i) geographical and (ii) state space. It is based on annual surface temperature fields computed over non-overlapping 100-yr segments (equilibrium simulations of Max Planck Institute ESM): last-millennium (800-1799), future climate projection (A1B scenario 2100-99), and 3100-yr unforced control. (i) In geographical space, a linear relationship between first and second moments is noted, most pronounced in the tropics: Negative (positive) regression slopes over the western (eastern) Pacific characterize the asymmetry of surface temperature ICV representing warm and cold ENSO extremes. In A1B the linear regressions largely retain their spatial structure but change in intensity and location. (ii) In state space spanned by a parsimonious set of dominating principal components of the moment fields, cluster centroids reveal the underlying asymmetry dynamics as ENSO and MODOKI modes. (iii) Extreme value estimates are affected by this first-second moment relation accounting for clustering of extremes.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2016
- Bibcode:
- 2016AGUFMNG11A..06Z
- Keywords:
-
- 3315 Data assimilation;
- ATMOSPHERIC PROCESSES