When we tackle large-scale spatial epidemiological models at the scale of interconnected cities, the choice of a modeling paradigm is not straight. In this paper, we present a hybrid approach (named MicMac) coupling Ordinary Differential Equations (ODE) Susceptible-Infected-Recovered (SIR) model (at the scale of cities) and agent-based transportation model (between cities). Beyond this innovative modeling approach, the main challenge investigated is related to the exploration of various disease mitigation strategies, such as quarantine or avoidance,  in this new modeling paradigm. Following the Geocomputation paradigm introduced by Openshaw during 80’s, we use some new evolutionary algorithms actually only supported by HPC (High Performance Computing) to explore a large combinatory of control strategy and finally try to cartography part of the complex behavior expressed by this coupled model. In particular, we use (a) the PSE exploration method, included inside the OpenMole exploration tool, to have an exhaustive picture of all the patterns we can find in the model and (b) the CP-Profile algorithm that allows the modeler to conduct a global sensitivity analysis eliciting the single parameters that can be balanced by the variation of all the other parameters. As an application, we explore the impact of some control strategies at global or individual scales, including quarantine, avoidance or risk culture. The aim is to identify benefits and drawbacks of each approach and thus to highlight the necessary complementary of both when simulating such large-scale.