Abstract
For 35 seasons in the years 1974–84, the importance of seasonal anomalies in tropical diabatic heating was investigated for the circulation in the tropics and in the extratropics. The heating was estimated from anomalies in outgoing longwave radiation as measured by satellite and was prescribed as a forcing in a linear study state model. With this model a small part of the observed spatial variance in the streamfunction anomalies in the tropics and lower midlatitudes in the northern hemispheric winter and autumn (5–10%) could be explained. In the tropics, in particular in the central Pacific at 700 mb, the explained variance was largest (10–25%). When the beating was exceptionally large, as during El Niño 1982–83, the similarity between observed and simulated streamfunction anomalies was much better than average not only directly over the major heat sources in the tropics but also in midlatitudes. In spite of the simplicity of the model and the neglect of the other forcing terms, the explained variance was between 25 and 60% in these regions.
