Current anthropogenic aerosol emissions are estimated to force a global mean cooling of 0.5 K. However, the climate response to aerosol forcing is highly spatially variable, giving a temperature much larger than 0.5 K in certain regions. While anthropogenic aerosol emissions have evolved historically and are likely to reduce rapidly in the coming decades in response to stricter air quality policies, the climate response to these emissions changes is not well understood because it depends on poorly constrained processes which have a range of representations in earth system models. I will discuss two projects that aim at improving understanding of the climate response to anthropogenic aerosol forcing using CanESM. First, I will discuss the development of a more detailed aerosol model for CanESM. I will present unphysical behaviour revealed by simulations with the initial version of the model, summarize model development that resulted to correct the unphysical behaviour, and show preliminary results from the improved version of the model. Second, I will outline the Regional Aerosol Model Intercomparison Project (RAMIP) simulations, in which one region rapidly reduces anthropogenic aerosol and precursor emissions while the rest of the world does not. I will show a preliminary analysis of how detectable the aerosol-forced response in available simulations is.