STUDY OF THE EQUATORIAL ATLANTIC OCEAN MIXING LAYER USING A ONE-DIMENSIONAL TURBULENCE MODEL

The General Ocean Turbulence Model (GOTM) is applied to the diagnostic turbulence field of the mixing layer (ML) over the equatorial region of the Atlantic Ocean. Two situations were investigated: rainy and dry seasons, defined, respectively, by the presence of the intertropical convergence zone and by its northward displacement. Simulations were carried out using data from a PIRATA buoy located on the equator at 23 degrees W to compute surface turbulent fluxes and from the NASA/GEWEX Surface Radiation Budget Project to close the surface radiation balance. A data assimilation scheme was used as a surrogate for the physical effects not present in the one-dimensional model. In the rainy season, results show that the ML is shallower due to the weaker surface stress and stronger stable stratification; the maximum ML depth reached during this season is around 15 m, with an averaged diurnal variation of 7 m depth. In the dry season, the stronger surface stress and the enhanced surface heat balance components enable higher mechanical production of turbulent kinetic energy and, at night, the buoyancy acts also enhancing turbulence in the first meters of depth, characterizing a deeper ML, reaching around 60 m and presenting an average diurnal variation of 30 m.

Regional representativity of AERONET observation sites during the biomass burning season in South America determined by correlation studies with MODIS Aerosol Optical Depth

This paper presents an analysis of ground-based Aerosol Optical Depth (AOD) observations by the Aerosol Robotic Network (AERONET) in South America from 2001 to 2007 in comparison with the satellite AOD product of Moderate Resolution Imaging Spectroradiometer (MODIS), aboard TERRA and AQUA satellites. Data of 12 observation sites were used with primary interest in AERONET sites located in or downwind of areas with high biomass burning activity and with measurements available for the full time range. Fires cause the predominant carbonaceous aerosol emission signal during the dry season in South America and are therefore a special focus of this study. Interannual and seasonal behavior of the observed AOD at different sites were investigated, showing clear differences between purely fire and urban influenced sites. An intercomparison of AERONET and MODIS AOD annual correlations revealed that neither an interannual long-term trend may be observed nor that correlations differ significantly owing to different overpass times of TERRA and AQUA. Individual anisotropic representativity areas for each AERONET site were derived by correlating daily AOD of each site for all years with available individual MODIS AOD pixels gridded to 1 degrees x 1 degrees. Results showed that for many sites a good AOD correlation (R(2) > 0.5) persists for large, often strongly anisotropic, areas. The climatological areas of common regional aerosol regimes often extend over several hundreds of kilometers, sometimes far across national boundaries. As a practical application, these strongly inhomogeneous and anisotropic areas of influence are being implemented in the tropospheric aerosol data assimilation system of the Coupled Chemistry-Aerosol-Tracer Transport Model coupled to the Brazilian Regional Atmospheric Modeling System (CCATT-BRAMS) at the Brazilian National Institute for Space Research (INPE). This new information promises an improved exploitation of local site sampling and, thus, chemical weather forecast.

The Diffusion Kernel Filter

A particle filter method is presented for the discrete-time filtering problem with nonlinear ItA ` stochastic ordinary differential equations (SODE) with additive noise supposed to be analytically integrable as a function of the underlying vector-Wiener process and time. The Diffusion Kernel Filter is arrived at by a parametrization of small noise-driven state fluctuations within branches of prediction and a local use of this parametrization in the Bootstrap Filter. The method applies for small noise and short prediction steps. With explicit numerical integrators, the operations count in the Diffusion Kernel Filter is shown to be smaller than in the Bootstrap Filter whenever the initial state for the prediction step has sufficiently few moments. The established parametrization is a dual-formula for the analysis of sensitivity to gaussian-initial perturbations and the analysis of sensitivity to noise-perturbations, in deterministic models, showing in particular how the stability of a deterministic dynamics is modeled by noise on short times and how the diffusion matrix of an SODE should be modeled (i.e. defined) for a gaussian-initial deterministic problem to be cast into an SODE problem. From it, a novel definition of prediction may be proposed that coincides with the deterministic path within the branch of prediction whose information entropy at the end of the prediction step is closest to the average information entropy over all branches. Tests are made with the Lorenz-63 equations, showing good results both for the filter and the definition of prediction.