Water ice clouds play a key role in the radiative transfer of the Martian atmosphere, impacting its thermal structure, its circulation, and, in turn, the water cycle. Recent studies including the radiative effects of clouds in global climate models (GCMs) have found that the corresponding feedbacks amplify the model defaults. In particular, it prevents models with simple microphysics from reproducing even the basic characteristics of the water cycle. Within that context, we propose a new implementation of the water cycle in GCMs, including a detailed cloud microphysics taking into account nucleation on dust particles, ice particle growth, and scavenging of dust particles due to the condensation of ice. We implement these new methods in the Laboratoire de Météorologie Dynamique GCM and find satisfying agreement with the Thermal Emission Spectrometer observations of both water vapor and cloud opacities, with a significant improvement when compared to GCMs taking into account radiative effects of water ice clouds without this implementation. However, a lack of water vapor in the tropics after Ls = 180° is persistent in simulations compared to observations, as a consequence of aphelion cloud radiative effects strengthening the Hadley cell. Our improvements also allow us to explore questions raised by recent observations of the Martian atmosphere. Supersaturation above the hygropause is predicted in line with Spectroscopy for Investigation of Characteristics of the Atmosphere of Mars observations. The model also suggests for the first time that the scavenging of dust by water ice clouds alone fails to fully account for the detached dust layers observed by the Mars Climate Sounder.