Forecasting El Nino and La Nina Using Spatially and Temporally Structured Predictors and a Convolutional Neural Network

El Nino-Southern Oscillation (ENSO) is characterized by large-scale fluctuations of sea surface temperature (SST) in the central and eastern tropical Pacific accompanied by changes in the atmospheric circulation. ENSO events are of two main types: El Nino and La Nina. Oceanic Nino index (ONI) determines the five consecutive three-month running mean of SST anomalies, in the Nino 3.4 region (5 degrees S-5 degrees N, 170 degrees W-120 degrees W). El Nino is a phenomenon in the equatorial Pacific Ocean characterized by a value of greater than 0.5 degrees C for ONI. La Nina is a phenomenon in the equatorial Pacific Ocean characterized by a value of less than -0.5 degrees C for ONI. The lingering of either of these two phenomena could induce severe droughts, whereas either of them following the other could cause massive floods. In both cases, deaths and substantial pecuniary loss are unavoidable, making their forecast of great significance. This study takes over the challenge of forecasting these two phenomena with one year lead time, which has proven difficult in the literature. This research's contribution is restructuring ENSO events' predictors, including SST, sea level pressure, surface wind speed, and wind stress in a spatially and temporally meaningful way and designing a convolutional neural network that takes advantage of this structure to forecast ENSO events of different types (i.e., Central Pacific and Eastern Pacific) in the next year. Not only a high precision in forecasts was achieved but also it was shown that the proposed model has the potential to achieve higher recalls if a larger number of samples from the positive class would become available.

Automated summary

ENSO, characterized by sea surface temperature fluctuations in the central and eastern tropical Pacific, has two main types: El Nino and La Nina. Forecasting these phenomena with high precision remains a challenge that requires constructing models effectively.