Despite tremendous increase in computing power, current climate models resolution (~100km) is still too low to explicitly represent several key processes in the atmosphere and ocean, which must therefore be parameterized and constitute one of the main causes of systematic errors in global ocean-atmosphere coupled models. Small-scale processes can play a major role in the variability of the climate at global scale through the intrinsic non-linearity of the system and the positive feedback associated with the ocean-atmosphere interactions. In this study high-resolution simulations of an extended tropical channel are performed in order to allow feedback of regional processes on the tropical climate mean state. In this perspective the atmospheric WRF model is coupled with the oceanic NEMO model via the OASIS3-MCT coupler. We analyze 20-years of coupled simulations with two different horizontal resolution, namely 0.75 and 0.25 degree. Comparisons between those simulations allow to determine the impacts of horizontal resolution on the climate mean state and its variability. The analysis focuses on the following key climate components: 1) the different monsoon systems and the associated regional circulations in the Indian, West African and North American regions, 2) the spatial and temporal distributions of the large-scale El-Niño Southern Oscillation, and 3) the upwelling systems in coastal and equatorial regions in the Pacific and Atlantic basins.