Satellite-based earth observation systems provide an effective means to detect and assess global changes and natural hazards. Recently developed numerous international and national networks, such as the GPM (Global Precipitation Mission), have global or regional coverages of various targets with a much finer time span than before. These aggressively progressed technologies in earth-monitoring have dramatically increased the volume and variety of spatio-temporal data. As a result, it has greatly boosted modelling and simulation of variations and events occurring irregularly in space and time with rate and magnitudes that are spatially heterogeneous and nonstationary in time. With an aim of gathering scientists to share their latest experiences and to explore new research opportunities in this field, our session invites presentations on integrating satellite remote sensing with other techniques, or any other modelling approaches, to model and evaluate changes and hazards in space and over time at global/regional scale. Our focus is on data, methods and applications where spatio-temporal aspects, with the inclusion of temporal covariates and the adaptation to irregularly occurring extremes in space, of extremes and disasters are explicitly taken into account.