Rainfall has been recognized as the main trigger of shallow landslides, and recent changes in rainfall patterns may have a significant impact on the frequency and amplitude of landslides, resulting in numerous fatalities and enormous economic losses. As a result, the landslide hazards are assessed by accounting for the temporal probability of a landslide by incorporating rainfall exceedance probability and landslide susceptibility into consideration. Research investigations have shown that hydrological records in some cases show non-stationarity in the form of increasing or decreasing trends and abrupt changes. Therefore, using the stationary condition to assess the temporal probability of landslides is no longer valid. In this study, the rainfall non-stationarity for the multiple sites of the Wayanad District of Kerala is taken into account for developing the landslide hazard map (LHM). The landslide susceptibility map (LSM) of Wayanad district is created in the first stage of landslide hazard mapping using the Random Forest (RF) model. The rainfall threshold exceedance probability is then calculated by utilizing a non-stationary extreme value analysis based on a novel multisite rainfall information technique, which takes into account antecedent rainfall for up to three days. The landslide susceptibility map and the temporal probability of rainfall are then combined to create the Wayanad district's LH maps for a 50 year return period. The preparation of LHM is found to be more resilient when the non-stationarity of the rainfall is taken into account.