Ocean Color and the Equatorial Annual Cycle in the Pacific

The presence of chlorophyll, colored dissolved organic matter (CDOM) and other scatterers in ocean surface waters affect the flux divergence of solar radiation and thus the vertical distribution of radiant heating of the ocean. While this may directly alter the local mixed-layer depth and temperature (Martin 1985; Strutton & Chavez 2004), non-local changes are propagated through advection (Manizza et al. 2005; Murtugudde et al. 2002; Nakamoto et al. 2001; Sweeny et al. 2005). In and coupled feedbacks (Lengaigne et al. 2007; Marzeion & Timmermann 2005). Anderson et al. (2007), Anderson et al. (2009) and Gnanadesikan & Anderson (2009) have performed a series of experiments with a fully coupled climate model which parameterizes the e-folding depth of solar irradiance in terms of surface chlorophyll-a concentration. The results have so far been discussed with respect to the climatic mean state and ENSO variability in the tropical Pacific. We extend the discussion here to the Pacific equatorial annual cycle. The focus of the coupled experiments has been the sensitivity of the coupled system to regional differences in chlorophyll concentration. While runs have been completed with realistic SeaWiFS-derived monthly composite chlorophyll ('green') and with a globally chlorophyll-free ocean ('blue'), the concentrations in two additional runs have been selectively set to zero in specific regions: the oligotrophic subtropical gyres ('gyre') in one case and the mesotrophic gyre margins ('margin') in the other. The annual cycle of ocean temperatures exhibits distinctly reduced amplitudes in the 'blue' and 'margin' experiments, and a slight reduction in 'gyre' (while ENSO variability almost vanishes in 'blue' and 'gyre', but amplifies in 'margin' - thus the frequently quoted inverse correlation between ENSO and annual amplitudes holds only for the 'green' / 'margin' comparison). It is well-known that on annual time scales, the anomalous divergence of surface currents and vertical upwelling acting on a mean temperature field contribute the largest term to SST variability (Köberle & Philander 1994; Li & Philander 1996). We examine whether it is changes in the surface currents (driven by the annual cycle of winds) or changes in the mean temperature fields (driven by enhanced penetration of solar radiation) that drive the differences between the coupled models. We do this using a simple linear equatorial-wave model, which, when forced with an annual harmonic of wind stresses, reproduces the essential characteristics of annual ocean current anomalies. The model solves the linearized Boussinesq equations by expansion into discrete modes in all spatial dimensions (McCreary 1981; Lighthill 1969). Both the wind forcing and the (laterally homogeneous) background density profile are constructed as approximations to the coupled model fields. The annual perturbation currents from the wave model are then used to advect the mean temperature fields from the coupled model experiments. While the difference in the mean stratification explains the difference between the 'green' and 'blue' cases. For the other two cases, it appears that changes in the annual wind fields need also be taken into account. An initial hypothesis is that the hemispheric asymmetry in the annual amplitude of wind stress curl that is most important in setting the amplitude of the annual cycle on the equator.