Patterns of urban development affect flows of material and energy within urban settlements and exchanges with their surroundings. In recent years the quantitative estimation of the components of the so-called urban metabolism has increasingly attracted the attention of researchers from different fields. To contribute to this effort we developed a modelling framework for estimating carbon exchanges in relation to alternative land-use scenarios. The framework bundles three components: (i) a Cellular Automata model for the simulation of the urban land-use dynamics; (ii) a transportation model for estimating the variation of the transportation network load and (iii) the ACASA (Advanced Canopy-Atmosphere-Soil Algorithm) model tightly coupled with the mesoscale weather forecasting model WRF. We present and discuss the results of an example application on the City of Florence.