The story begins in February 2002 when, in the bustling halls of the International Research Institute for Climate and Society (IRI), a team of researchers began issuing forecasts based on what has now become the world’s longest archive of real-time monthly El Niño Southern Oscillation (ENSO) forecasts.
Fast forward to 2023, and this ENSO archive has grown into a 21-year-long treasure trove of 253 real-time forecasts. Analyzing these forecasts was like opening a time capsule, revealing snapshots of past efforts to predict ENSO.
As researchers dug into the data, patterns emerged. They noticed that both dynamical and statistical models struggled with long-lead forecasts of ENSO. However, they discovered that dynamical models had a secret weapon – they performed significantly better than statistical models through the “spring predictability barrier,” which has long been the bane of ENSO forecasters because it is the season when ENSO predictions are least skillful.
But the most intriguing plot twist came when the team examined how well the models predicted the onset of El Niño and La Niña events. It turns out the forecaster’s crystal ball is clearer when El Niño emerges than for La Niña. This asymmetry in predictability adds a new layer of complexity to our understanding of ENSO dynamics.
The researcher’s journey through this forecast archive revealed that predicting ENSO is not unlike predicting the plot of a complex novel – some twists are easier to anticipate than others, and the accuracy often depends on both the storyteller (the model) and the particular chapter (the specific ENSO event).
As the team concluded their two-decade journey through ENSO forecasts, they gained valuable insights for future predictions, underscoring that in climate science, understanding the past is crucial for anticipating the future. This analysis, requiring over two-decades of patient data collection, showcases IRI’s dedication to ensuring long-term research efforts are used in real world decision making and risk assessments. This achievement stems from the continuous support of numerous forecasting centers around the globe (currently 28 in number) that generously share their forecast data publicly through the IRI Data Library, and the visionary efforts of past eminent IRI scientists who initiated this collection, highlighting the collaborative nature of climate science.
Disclaimer: The scientific views or opinions expressed herein are those of the author(s) and do not necessarily reflect the views of IRI, CCSR of Columbia University, or NWS, NOAA, or the Department of Commerce.