Abstract Individual-based models, which permit a fine description of the dynamics of a population are costly in computation time and need many simulations for outlining global laws. Conversely aggregate models take less time for simulation and provide analytical understanding of the global evolution. Therefore they can sometimes profitably replace individual-based models. However, in many cases these models are linked to a model of environment which evolves over time. In this paper we show that the relevance of using an aggregate vs. an individual based model of water consumption highly depends on the information available on the resource. We use two versions of a consumption model (individual-based and aggregate), simulating a reversible diffusion process which depends on the information available on the resource. We compare the results of the two models for various kinds links between the resource and consumption models and corresponding to the information about the resource provided to the consumers. We propose and test different properties of this information which might lead to difference of the results.