Mineral dust interacts with the climate by modifying the Earth's radiation budget and by transporting nutrients to the terrestrial and marine ecosystems. To estimate how the atmospheric dust loading will change in the future it is important to understand the processes that control the quantity of dust in the atmosphere. Current dust cycle models are unable to predict changes in the extent of arid and semi-arid regions caused by the transient response of vegetation cover to the climate. As a consequence, it is not possible to predict the expansion and contraction of these regions on seasonal and inter-annual time scales. A new dust cycle model is developed which uses the Lund-Potsdam-Jena dynamic global vegetation model to calculate time evolving dust sources. Surface emissions are calculated by simulating the processes of saltation and sandblasting. Dust is transported in the atmosphere by advection, convection and diffusion and removed from the atmosphere by dry deposition and sub-cloud scavenging. To improve the performance of the model, threshold values for vegetation cover, soil moisture, snow depth and threshold friction velocity, used to determine surface emissions are tuned. The effectiveness of three subcloud scavenging schemes are also tested. The tuning experiments are evaluated against multiple measurement datasets. The ability of the new model to predict seasonality in the dust cycle is evaluated. The model is successful at predicting the seasonality in dust emissions from North Africa, South Africa, Patagonia, North America, and Asia but not in Australia where LPJ is unable to predict the vegetation dynamics correctly. In all regions maximum emissions occur when low precipitation combines with a high frequency of wind speed events greater than 2ms-1. In Patagonia, surface emissions are strongly anti-correlated with precipitation because wind speeds exceed 2ms-1 continuously throughout the year. Vegetation cover constrains emissions in North America, Central Asia, Eastern China and South Africa. The new model has been used to investigate whether changes in vegetation cover in the Sahel can explain the four-fold increase in dust concentrations measured at Barbados during the 1980s relative to the 1960s. Results showed there was an expansion of the Sahara in 1984 relative to 1966 resulting in a doubling of emissions from the Sahel. This alone is not enough to account for the high dust concentrations in 1984. This finding adds strength to the hypothesis that human induced soil degradation in North Africa may be responsible for the increase in high dust concentrations at Barbados during the 1980s relative to the 1960s. To predict how dust source areas will change in the future it is important to disentangle processes which cause natural variability from anthropogenic effects.