Prediction of tropical Atlantic sea surface temperature

Abstract

The variability of sea surface temperature (SST) in the tropical Atlantic Ocean has a significant influence on the rainfall over Northeast Brazil and the Sahel. The predictability of the SST fluctuations in this region has, therefore, been a central concern for Atlantic climate studies. Here we show that the north tropical Atlantic SST variability has significant predictability using statistical models at short lead times (≤1 year), and using a coupled ocean-atmosphere model for longer-range forecasts (lead times up to about 3 years). While the remote influence of the IndoPacific, primarily due to El Nino, has been found to enhance the predictability of SST in the north Atlantic at short lead times, local ocean-atmosphere interactions dominate the predictable dynamics farther south, and at longer time scales.