Extended Abstract (536K)14C.5
Use of Granger causalities to examine air-sea feedbacks in the tropical equatorial Pacific Ocean
Brian Getzewich, Aerospace Corporation, El Segundo, CA; and C. A. Clayson
Air-sea interactions contain a high degree of non-linear feedbacks in the tropical Pacific Ocean, particularly in relation to sea surface temperature. Some processes, such as boundary layer responses to surface forcings, can respond within hours and are confined to small regions. On the other hand, changes in sea surface temperature throughout the tropical Pacific can potentially bring the onset of atmospheric circulatory features that persist from several months to years and stretch across a quarter to half of the globe. In this study we examine feedbacks related to sea surface temperature regulation on time scales from hours to days.
In order to examine the importance of the surface energy budget components on causing variability in sea surface temperature, a non-conventional time series analysis technique originally developed for economic applications was adapted. Granger causality theory was developed initially as a means to examine variable relationships in the field of economics. Unlike traditional correlation techniques, this methodology is used to detect feedback relationships within the data between the processes that generated them. In other words, this method determines whether the ‘dependant variable’ is meaningfully responsive to changes in the ‘independent’ variable. The data used in this study was hourly data was extracted from the TAO buoys spread throughout differing regimes in the equatorial tropical Pacific. We examine the importance of feedback effects on hourly scales for different years and different locations across the Pacific. The importance of the surface energy budget fluxes in determining sea surface temperature was also evaluated for several locations and on an interannual basis. The hourly analysis demonstrated that causality and feedbacks were similar from year to year and from location to location. The use of daily averages however showed large differences in the importance of such parameters as solar radiation and latent heat flux from year to year. Selected results from this research will be shown.
Session 14C, Ocean-Atmosphere Interaction III (Parallel with Sessions 14A, 14B, and 14D)
Thursday, 2 May 2002, 2:00 PM-3:30 PM
Previous paper Next paper