A simple dynamic model is developed to describe the observed interactions between the atmosphere and the soil moisture patterns of the northern Sahel. In the model, the atmosphere follows quasi-geostrophic dynamics, while landatmosphere coupling is described by simple linear relationships. Dry surfaces heat the atmospheric boundary layer, while wet surfaces cool the boundary layer, relative to the equilibrium state of the atmosphere and land surface. In turn, cloud processes, which are assumed to maximise in the cool, humid phase of an atmospheric disturbance, cool the land surface through wetting (rainfall) and reduction of the incoming solar flux. These assumptions lead to a linear system which can be solved numerically to obtain modal solutions, and the adjoints (optimal excitation) of these. Moist convective influences on the atmospheric state are not explored in detail. The coupling with the land surface leads to the existence of unstable modes, which do not exist in the atmosphere-only part of the system. Solutions can be easterly or westerly propagating, according to wave number, with the longer waves tending to be easterly. Propagation relies on a favourable configuration between the atmospheric and soil moisture anomalies: easterly propagation requires the surface temperature pattern to be shifted to the east of the atmospheric temperature pattern. In contrast, optimal excitation of the fastest-growing mode occurs when the atmospheric pattern has a thermal anomaly lying to the east of a strong surface temperature (and moisture) anomaly. These results have value for weather prediction, and indicate the usefulness of soil moisture data for forecasters.