Intercomparison between observed and simulated variability in global ocean heat content using empirical mode decomposition, part I: modulated annual cycle

This study proposes a new more precise and detailed method to examine the performance of IPCC AR4 models in simulation of nonlinear variability of global ocean heat content (OHC) on the annual time scale during 1950-1999. The method is based on the intercomparison of modulated annual cycle (MAC) of OHC and its instantaneous frequency (IF), derived by Empirical Mode Decomposition and Hilbert-Huang Transformation. In addition to indicate the general agreement in gross features globally between models and observation, our results point out the problems both in observation and in modeling. In the well observed Northern Hemisphere, models exhibit extremely good skills to capture nonlinear annual variability of OHC. The simulated MACs are highly correlated with observations (> 0.95) and the IF of MACs varies coherently with each other. However, in sparsely observed Southern Hemisphere (SH), even though the simulated MACs highly correlate with observations, the IF shows significant difference. This comparisons show that the models exhibit coherent variability of IF of MACs in SH with each other, but not with observations, revealing the problems in the objective analyzed dataset using sparse observations. In the well observed tropic region, the models lack the coherence with the observations, indicating inadequate physics of the models in the tropical area. These results illustrate that the proposed method can be used routinely to identify problems in both models and in observation of the global ocean as a critical component of global climate change.