Anthropogenic land use and land cover change (LULCC) influence global rates of wind erosion and dust emission, yet our understanding of the magnitude of the responses remains poor. Analyses of LULCC and land management-aeolian process interactions require field measurements and models that are sensitive to human and ecological drivers and provide an acceptable level of certainty in change detection. However, current approaches to measuring and modelling wind erosion and dust emission are typically highly uncertain as they inadequately account for the large spatial and temporal variability in sediment transport. Field measurements need to be sufficiently robust to establish that aeolian sediment transport responses to change are statistically significant at a desired confidence level and capture the effects of boundary-layer interactions. Here, we demonstrate that small sample sizes are often inadequate for monitoring aeolian processes and quantifying anthropogenic interactions. The spatial variance in transport rates may be many times larger than the temporal variance depending on the land surface aerodynamics and sediment supply, which are influenced by land cover and land management. Statistical rigour and the straightforward application of a sampling design can reduce the uncertainty and detect change in sediment transport over time in response to management and between land use and land cover types. We discuss measurement uncertainty effects on dust model calibration and applications to quantify anthropogenic interactions. Model uncertainty has arisen from a lack of sensitivity to spatial variability in land surface controls (e.g., vegetation), and the uncertainty in measurements used for calibration and testing. We examine how robust field measurements can be combined with new modelling approaches that resolve the spatial variability in transport controls to improve dust model sensitivity and reduce uncertainty in assessments of the impacts of LULCC.