Statistics of coupled ocean and atmosphere intraseasonal/seasonal anomalies in Reanalysis and AMIP data and implications on the seasonal forecast predictability

Locally coupled persistent anomalies in the global SST and lower troposphere vorticity with time scales ranging from a week to a season are examined. The objective is to document the relative importance, and seasonal dependency, of ocean-driving versus atmosphere-driving scenarios in the coupled anomalies with different time scales at different geographical locations. An empirical CISK-like air-sea interaction rule has been applied to detect whether the forcing mechanism comes primarily from the ocean or from the atmosphere in the coupled anomalies. Effectively, we examine the coupled anomalies on a case-by-case basis to separate the statistics of the coupled anomalies with one forcing direction from the other. The results are further ratified with an independent method which examines the temporal phase relation between the atmospheric and oceanic anomalies. In additional, we have checked the consistency of the case-by-case statistics with the traditional cross correlation between the anomalies.

The results obtained from the NCEP/NCAR Reanalysis data clearly suggest that in general the coupled persistent anomalies originate more frequently from the ocean in the tropics and from the atmosphere in the middle latitudes. Over the eastern Pacific subtropic oceans, atmosphere-driving scenarios are more dominant compared to other parts of oceans. On the other hand, the number of ocean-driving cases is comparable to that of atmosphere-driving cases over the warm pool. The same procedures have been applied to the NCEP-Atmospheric Model Intercomparison Project (AMIP) data. The results from the AMIP data show a significant bias towards ocean-driving scenario in most of the middle latitude regions. The AMIP data also show a number of cases in which anticyclonic vorticity over a warm SST is spuriously larger than that revealed from the reanalysis data. We will discuss the implications of these results in the currently operational extended forecast procedures where the SST is regarded as the primary forcing for the future atmospheric anomalies.