Experiments on the effect of tropical Atlantic heating anomalies upon GCM rain forecasts over the Americas

Abstract

A series of real data experiments is performed with a general circulation model in order to ascertain the sensitivity of extended range rain forecasts over the Americas to the structure and magnitude of tropical heating anomalies. The emphasis is upon heat inputs over the tropical Atlantic which have shown particularly significant drying influences over North America in our prior simulations. The heating imposed in the prior experiments is compared to the condensation heating rates that naturally occur in the forecast model, and shown to be excessive by approximately a factor of two. Present experiments reduce the imposed anomaly by a factor of three, and also incorporate sea-surface temperature decreases over the eastern tropical Pacific Ocean. The new experimental results are in many ways consistent with our prior results. The dry North American response is statistically more significant than the South American response, and occurs at least as frequently in the different members of the experimental ensembles as in our prior experiments. The drying effect is accentuated by the presence of East Pacific cooling, but this does not appear to be the dominant influence. Over tropical South America, the Pacific and Atlantic modifications produce compensating influences, with the former dominating dominant, and allowing increased rainfall over the Amazon Basin.