Context. The Martian methane and water cycle are subject of ongoing research through simulation. Exchange with the subsurface has a potentially strong impact, but is often neglected.

Aims. For methane, I determine if adsorption with an increased enthalpy can explain the observed seasonal variations and conflicting observations by the Trace Gas Orbiter and the Curiosity rover. For water, the impact of adsorption and ice formation in the subsurface on the global cycle is studied. A new way of initializing the soil, by running a decoupled subsurface model, is tested. Depths of stable subsurface ice and subsurface water distributions are studied.

Methods. A General Circulation Model (GCM) is used with a purely diffusive subsurface model. For methane, different initial states, source scenarios, and decay times are tested. For water, a model without an active atmosphere is implemented to provide an initial state. The effect of the subsurface with this initial state on the full atmospheric water cycle is tested.

Results. For methane, a strong influence on the global methane cycle is observed. Seasonal variations measured at Gale Crater are reproduced, but the conflicting observations cannot be explained by adsorption. For water, the new initialization can be used without completely disrupting the water cycle. It leads to a generally wetter atmosphere, in conflict with observations. Found ice table depths do not match well with observations, but ice profiles reproduce previous findings.

Conclusion. Methane adsorption is able to partly explain observed variations, but cannot be the only process to influence methane abundances. The new initialization method for water works well in principle, but a more refined model is needed for more realistic results.