A network of connected wetlands supports migratory movements of waterfowl. These networks are rapidly changing due to intensive human activities around natural habitats. Quantifying how anthropogenic factors change waterfowl movements via a reduction of habitat availability and quality can facilitate a better understanding of the dynamics of these migration networks, and provide early-warning signals for network collapse. Using satellite tracking data for greater white-fronted geese (Anser albifrons) in the East Asian-Australasian Flyway, we tested how environmental factors (i.e., anthropogenic and ecological factors) influence geese movement patterns at stopover sites. We found that these factors, e.g., percentage of farmlands in the landscape, and proximity index of wetland patches, accurately predicted percentage of flying time and the median movement distance of tracked geese at stopover sites. Farmlands may increase energy consumptions in stopover sites because the geese flew more frequently, made longer movements, and switched their behaviour more frequently in landscapes with a higher proportion of farmlands. Goose movements were constrained in natural habitats, as a higher proportion of water and wetlands increased their movements, and thereby increased flying time and median movement distances. We suggest that using environmental factors monitored by remote sensing techniques to predict bird movement patterns is a powerful quantitative tool to measure quality of stopover sites. The changes in environmental factors in these stopover sites can be used as an indicator for the probability of losing a site from a migration network, and thereby generates insights for setting priorities in conservation planning of migratory birds.