The paper discusses problems encountered while implementing a downscaling experiment for testing the potential impacts of changes in temporal precipitation patterns on the biogeochemical properties of the lagoon of Venice. The paper shows how coupling different models implies redefining the boundaries, as well as space and time scales of the systems to be modelled, carefully revising model assumptions and parameterization, and possibly integrating new elements in the model. The paper also details how, in order to account for changes and adaptation in planktonic pools, a structural dynamic model was implemented by linking the kinetic parameters of a pre-existing calibrated functional response to the recent evolution of environmental conditions. Parameterizations of the boundary conditions of rescaled models and of links among models are also discussed.