Abstract
A new general circulation model, developed to run on a coarse grid (8 × 10° resolution) at the Goddard Institute for Space Studies is employed to investigate the potential use of ground moisture anomalies for seasonal climate prediction. For three different summertime simulations, the ground moisture on 1 June over the United States is reduced to ¼ of its value in the control run. The results show that the subsequent surface air temperature is significantly higher throughout most of the summer, while the precipitation decreases, especially in June and July. Knowledge of late spring ground moisture anomalies should thus be an aid in predicting summertime climate.